USPTO Patents Application 10681421

P50438-1

Ma7I™R0MBOTIC

JklSEHTS

gTRU OF '™« TUVBSHm

This invention relates to monoclonal antibodies
,^s) that bind to a human coagulation -
cofactor and their use as self -limiting rnhrbxtors of
throKtoosls in combination with plasminogen activators.

„,.^^^rs,Tm^ OF TP° TTOVEMTION

under normal circumstances, an injury, be it mxnor
or major, to vascular endothelial cells lining a blood
Vessel triggers a hemostatic response through a se<^ance
of events coi^only referred to as the coagulatron
..cascade.- The cascade culminates in the conversion of
sollle fibrinogen to insoluble fibrin which, -9-*^-
Tith platelets, forms a localized clot or thro::tous which
prevents extravasation of blood components. Wound
Laling can then occur followed by clot dissolution and
restoration of blood vessel integrity and flow.

The events which occur between injury and clot
f oration are a carefully regulated and linked series of
reactions. In brief, a nu-toer of plasma coagulation
proteins in inactive proenzyme forms and cofactors
circulate in the blood. Active enzyme complexes are
assembled at an injury site and are sequentially
activated to serine proteases, with each successive
serine protease catalyzing the subse^ent proenzyme to
;rotease activation. This enzymatic cascade resul s in
each step magnifying the effect of the ^^^^
For an overview of the coagulation cascade see the first
chapter of ..Ttoombosis and Hemorrhage... J. Loscalzo and
A, schafer. eds., Blackwell scientific Publications,

Oxford, England (1994). M ^,,d

While efficient clotting limits the loss of blood
at an injury site, inappropriate formation of ^^o^ -
veins or arteries is a con^on cause of disability and

1

P50438-1

formation of clots on ^-^^^^f ^//J/,,,,, j^^art
organs, shunts and prostheses such as artxfrcxa

parenteral y . ^ • „„ „f the thrombin inhibitor,

of clotting by activation of the th
antithrombin III and inactivation of all

factors

However, due to their potency, heparin and L«
patients receiving continuous

vivo clotting txmes to be J" the

contributes substantially to the cost of
patient's inconvenience. ^^^.^^^
Further, the therapeutic target rang

. ^ 1 .oi of efficacy without placing the

P50438-1

and at concentrations
exceeds --- -- ana

..eater than 4 "^'^^' fJ^/ZJ^,,, taken to Iceep the

:rtre:r;. .rrrt^n. «... t. t„t.

range. ^^,,iant with slower and

Another approved anticoagulant wxt

• effect is warfarin, a coumarxn
longer lastxng effect xs ^^.^^ ^^.^ vitaxain K

derivative. Warfarxn acts by ^ prothrombin
dependent Post-translational .od f x^c^^^^^

and other Vitamin ^-^^P^^^^^'^/^;';,,,!, which blood

=1 oattern of anticoagulant actxon,
r;e:aerernon-c.otta..e at =oncent«-ons on^^^^^

-----T.rr.:rrrrLpa.n.
"-^Tar/rcLa.. ----^--ra^;::a

sustained reperfus.on of the .n „

rxrr:"r.T.-«.:.-.".-
.«»""-rr^:irrrr..r"."„ ...... -

unfractionated heparin, lo ^^^elet agents such

airect throK^in inhibitors or ^^^^^^^^^^

■ • T^latelet glycoprotexn IlD/xxx

as asp.r.n or platel g ^^^^^^ ^^.^

see Topol, ^ Heart J, 136, observation
combination of therapxes is base

..at Clot formation and a--^;-- "^^f^^^^^ ,..,,„ue
processes and thrombin ^^^Z::^; and during

a£ter the formation °* J <,,„ger et al, J

„d after ---/--V;,. 7, LsB, .
- coll Cardiol 31, «J ^^^^^^^^

...rers^e^ra:auabieage.s.d.^^^^^^^^^^

— — :"f:^f^^^

P50438-1

York (1997)) s'^d thrombin activity e=chibits a

:::::;;rJr:::e'.oiio.in. cessation o. ™

„,.b an ob.er.ea increase : 1,,..

hours following discontinuation of heparin.

et a . Cat.e.erizatio. and Cardiovascular

^4 25;-264 ,1998, and Granger, Circulation, 91 1929

19;5 ,1995) . Further, antiplatelet agents may be

■ A y.^r -bleeding or thrombocytopenia,
accompanxed by bleedxng ^^^^ ^.^^

Also, numerous clxnical trxais nav

aoses of thror^olytic agents lead to significant

alteration in plasma hemostatic markers. See Rao et

. J Clin invest, 101, 10-14 ,1988); Bovill et al.^

- T\ rsr-sfi — rir^ea^i:

rceratirif .aa to —ed clot d^solution,

the alteration in these hemostatic markers mirrors
increased liabilities of thror^olytic therapy,
;Lticularly the incidence of severe ble^ing^

Clearly, a need exists for antithrombotic agents
efficacious in controlling thro:*otic disorders while
maintaining hemostatic functions.

fn»«.1.Y OF THVEHTIOM

accordingly, one aspect of the present invention is
a method for inhibiting thrombosis in an animal
LIprising administering an effective dose of an anti-
rolgulation factor monoclonal antibody having s If-
limlting neutralizing activity in combination with

plasminogen activator. „,,rhod of

Another aspect of the invention is a method of
reducing a retired dose of a thrombolytic agent in
treatment of thrombosis in an ,
administering an anticoagulant ^^-^^^^^^^^^^J'^f ^'^^
component of the intrinsic coagulation pathway m
combination with the thrombolytic agent.

Another aspect of the invention is a method for
reducing a recjuired dose of a thro^olytic agent

4

P50438-1

treatment of thrombosis in an animal comprisxng
administering an anti-Factor IX monoclonal ant.body xn
combination with the thrombolytic agent

pT,Tr.iP T^TTfirHTPTION »^ '^"^ DRAWINGS

Figure 1 is a graph of experimental results
demonstrating the titration of normal human plasma wxth
the murine anti-Factor IX mAbs BCl and BC2 .

Figure 2 is a graph of experimental results
demonstrating the titration of normal human P^^-- -^^^
the murine anti-Factor IX mAbs 9E4(2)F4 and 11G4{1)B9.

Figure 3 is a graph of experimental results
demonstrating the titration of normal human plasma wxth
the murine anti-Factor X mAbs HFXHC and HFXLC and the
murine anti-Factor XI mAb HFXI.

Figure 4 is a histogram of experimental results
demonstrating the effect of heparin, acetylsalicylxc
acid and murine Factor IX mabs on activated partial
thromboplastin time (aPTT) at 60 minutes m a rat
carotid thrombosis model.

Figure 5 is a histogram of experimental results
demonstrating the effect of heparin, acetylsalicylxc
acid and murine Factor IX mabs on prothrombin txme at 60
minutes in a rat carotid thrombosis model.

Figure 6 is a histogram of experimental results
demonstrating the effect of heparin, acetylsalicylxc
acid and murine Factor IX mabs on occlusion of carotxd
artery flow in a rat carotid thrombosis model.

Figure 7 is a histogram of experimental results
demonstrating the effect of heparin, acetylsalicylxc
acid and murine Factor IX mabs on thrombus weight xn a
rat carotid thrombosis model.

Figure 8 is a histogram of experimental results
demonstrating the effect of heparin, the murine Factor
IX mab BC2, a chimeric Factor IX mab and humanxzed

5

P50438-1

factor IX m^s on aPTT at 60 minutes in a rat carotid

thrombosis model.

Figure 9 is a histogram of experimental results
demonstrating the effect of heparin, the -"-/^"^
ZX raab BC2 , a chimeric Factor IX ^ and hu.anr.ea
factor IX mAbs on throrc^us weight rn a rat carotxd
thrombosis model.

Figure 10 is a histogram of experimental results
demonstrating the effect of anti-Factor IX mab and
heparin on tPA-mediated reperfusion.

Figure II is a histogram of experimental results
demonstrating the effect of anti-Factor IX mab and
heparin on carotid vessel patency.

Figure 12 is a histogram of experimental results
demonstating the effect of anti-Factor IX mab and
heparin on time to restoration of blood flow.

Figure 13 demonstrates the effect of tPA on the
hemostatic parameters, fibrinogen, plasminogen and

^"'''Frg:r"e\4 demonstrates the effect of tPA, heparin
and anti-Factor IX mab on aPTT.

r.i.n.i.TT.ir.n DEF^^-^^t-rnw OF THF. INVENTION

All publications, including but not limited to
patents and patent applications, cited xn the
fpecff ication are herein incorporated by reference as
though fully set forth.

The present invention provides a varrety of
antibodies, altered antibodies and fragments thereof
directed against coagulation factors, «hrch are
Characterized by self-limiting

preferably, the coagulation factor rs from ^he rntrrnsrc
or coKo^on coagulation pathway. Most P-^"^^^;;*^^
anti-coagulation factor antibodies are ^'^---"^J^'
anti-Factor IXa, anti-Factor X,

Factor XI, anti-Factor XIa, anti-Factor VIII. antr

6

P50438-1

Pactorvxxxa, anti-.actor V, anti-.actor Va ant.-.actor
VII, anti-Factor Vila, anti-thron0.in or antx-
prothrornJ^in. Particnlarly preferred are ant.-Factor
antibodies. Exemplary anti-coagulation ^-"-^

. ■v,„,^ies are the humanized monoclonal antibodxes SB

SB 257;32 directed against human Factor XX, the chrmer.c
monoclonal antibody chaFIK directed

IX the murine monoclonal antibodies BCX, BC2 , 9^*'2)^4
and 1X04,1,B9 which are directed against ^ ° ^

.H/or Factor XXa or the murine monoclonal antibodies

ain:xx which are directed against human Factors

X and XI, respectively. Particularly P- ^"^^
anti-human Factor IX monoclonal antibody SB 2494X7^

The antibodies of the present invention can be
prepared by conventional hybridoma techniques phage
display coLinatorial libraries, i^unoglobulin chain
t^ffUng and humani.ation techniques to generate novel
self-limiting neutralizing antibodies. Also inc uded
Ire fully human having self-limiting — »

activity These products are useful in '^--P-"=
Tharmaceutical compositions for thrombotic and embolic
disorders associated with myocardial infarction,
!: table angina, atrial fibrillation, strojce, renal
damage, pulmonary e.*olism, deep vein thrombosis,
percutaneous translumenal coronary angioplasty,
disseminated intravascular coagulation, sepsis,
artificial organs, shunts or prostheses.

AS used herein, the term "self -limiting

► refers to the activity of an

neutralizing activity relers t

antibody that binds to a human coagulation factor,
preferably from the intrinsic and common pathways,
rduding^actor XX/XXa, X/Xa, Xl/XXa VXIX/VXIIa a d
v/va VXX/VIIa and thrombin/prothrombin and inhibits
:hro;a.osis in a manner such that limited modulation of
coagulation is produced. "Limited modulation of

P50438-1

coagulation" i. defined as an

as measured by prolongation of the activated partial
thromboplastin time (aPTT) , where plasma remains
clottable with aPTT reaching a maximal value de-P^*^^
increasing concentrations of monoclonal antibody Th.s
limited modulation of coagulation is in contrast to
plasma being rendered unclottable and exhibitrng an
infinite aPTT in the presence of increasing
concentrations of heparin. Preferably, the maxrmal aP^
value of the methods of the invention are wxthrn the
heparin therapeutic range. Most preferably, maxrmal
aPTT is within the range of about 3 5 seconds to about
100 seconds which corresponds to about 1.5 times to
about 3.5 times the normal control aPTT value^ In one
embodiment of the invention, aPTT is prolonged without
significant prolongation of prothrombin time (PT) .

The phrase "in combination with" refers to
administration of one therapeutic agent before, after or
concurrent with the administration of another
therapeutic agent in a single course of treatment^

"Altered antibody" refers to a protein encoded by
an altered in^unoglobulin coding region, which may be
obtained by expression in a selected host cell. Such
altered antibodies are engineered antibodies (e.g..
chimeric or humanized antibodies) or antibody fragments
lacking all or part of an immunoglobulin
region, e.g., Fv, Fab, Fab- or F(ab')2 and the liKa^

"Altered immunoglobulin coding region" refers
nucleic acid secjience encoding an altered antibody of
the invention. When the altered antibody is a CDR-
.rafted or humanized antibody, the sequences that encode
the complementarity determining regions (CDRs) from a
non-human immunoglobulin are inserted into a first
immunoglobulin partner comprising human variable
framework sequences. Optionally, the first

8

P50438-1

i^unoglobuUn partner is operatively linked to a second

immunoglobulin partner. „„^lslc
..First immunoglobulin partner" refers to a nuclerc
acid secjuence encoding a human framework or human
i:™unoglobulin variable region in which the -
naturally-occurring, CDK-encoding -^^^ ^^^^f
by the CDR-encoding regions of a donor antxbody
hLan variable region can be an immunoglobulrn heavy
Chain, a light chain (or both chains,, an analog or
functional fragments thereof. Such CDR regxons, located
within the variable region of antibodies
(immunoglobulins, can be determined by Known methods rn
the art. For example Kabat et al . in "Se^ences of
proteins of Immunological Interest-, 4th Ed., U.S.
Department of Health and Human Services, ^^^^^^f
institutes of Health (1987, disclose rules for locating
CDRS in addition, computer programs are known whxch
are useful for identifying CDR regions/structures.

■■second immunoglobulin partner- refers to another
nucleotide se^ence encoding a protein or
which the first immunoglobulin partner xs fused .n frame
L by means of an optional conventional linker sequence
(i e , operatively linked, . Preferably, it is an
Lunoglobulin gene. The second -'^-^f^/^fj;:
may include a nucleic acid sequence encoding the entire
constant region for the same (I.e., homologous, where
the first a!d second altered antibodies are derived from
the same source, or an additional ,i e.,

antibody of interest. It may be an — ^"^^^^"^f^^^
Chain or light chain (or both chains as part of a single
polypeptide, . The second immunoglobulin partner is not
limfted to a particular immunoglobulin class or isotype.
in addition, the second immunoglobulin partner may
comprise part of an immunoglobulin constant region such
as found in a Fab, or F(ab„ (i.e., a discrete part of
an appropriate human constant region or framework

9

P50438-1

region) . Such second Inununoglobulin partner may also

comprise a sequence encoding an integral

proLin exposed on the outer surface of a host cell

e g , as part of a phage display library, or a sequence

encoding a protein for analytical or diagnostic

detection, e.g.. horseradish peroxidase, P-

^"^^rfe::; :::-.c, ... .a. .a^. or ..a.,, are used
with their standard meanings. See, e.g., Harlow et al .
in "Antibodies: A Laboratory Manual", Cold Sprrng

Harbor Laboratory, (1988). ^^^^ribes
AS used herein, an "engineered antibody describes
a type of altered antibody, i.e., a full-length
synthetic antibody (e.g., a chimeric or h-anrzed
al^ibody as opposed to an antibody fragment) in which a
Tortion Of the light and/or heavy chain variable domains
of a selected acceptor antibody are replaced by
analogous parts from one or more donor antibodies which

Uificity for the selected epitope.
such molecules may include antibodies characterized by a
humanized heavy chain associated with an --"^^^"^
light chain (or chimeric light chain) , or vice versa.
Engineered antibodies may also be characterized by
alteration of the nucleic acid se^ences encoding the
acceptor antibody light and/or heavy variable domain
framework regions in order to retain donor antibody
binding specificity. These antibodies call
replacement of one or more CDRs (preferably all from
the acceptor antibody with CDRs from a donor antibody
described herein.

A "Chimeric antibody" refers to a type of
engineered antibody which contains a naturally-occurring
Triable region (light chain and heavy ^--^"^
from a donor antibody in association with light and
heavy chain constant regions derived from an acceptor
antibody.

P50438-1

A ..huraanized antibody- refers to - type of
engineered antibody having its CDRs derived fro. a non
hun,an donor iwaunoglobulin, the remaining
i^aunoglobuiin-derived parts of the - ^^^^^
derived from one or more human x^nunoglobul.ns In
addition, frame^or. s.pport id.es may b^^^^^^^^^^^^^

ro:rt;rarsrr"a. ro:..^. ..s., ^odgson

at al Bio/Technology, S, 421 (1991) ■ , ,

T;e term "donor antibody- refers to a monoclonal or
recoIinaL antibody which contributes the nucleic ac.d
se^ences of its variable regions CPKs or other
functional fragments or ^l^^;- TZ..
immunoglobulin partner, so as to provi
lunoglobuXin coding region and --"^-^--^^f
altered antibody with the antigenic spec.fic.ty
rtralizing activity ufeTn this

antibody. One donor -^^-::^ZZZ:.
invention is a mur.ne ^^""^^^^^^"^ ^^^^^ 3,i«ble
monoclonal antibody designated as ^« '
donor antibodies include the -"-/^"''^"^'^f 3,,
neutralizing monoclonal antibodies designated as BCl.
9.4(2,.4, 11G4(1)B9, ^F^LC and HPXX^ ^^^^^^^^
The term -acceptor antibody reters t
.ecor^inant antibodies heterologous to the donor

partner. Preferably, a human antibody xs the accept
antibody.

"CDRs- are defined as the complementarity
determfnLg region amino acid sec^ences of an antibody
::::: are the hypervanable regions of i:™^og obulin
,eavy and light chains. ' ."^^f 'J /j,!

■■sequences of Proteins of Immunological Interest ,

P50438-1

H(=alth and Human Services,
Fd US. Department of Heaicn duv^ ^vv-oo
f H^^lth (1987). There are three
national ----^ ^ ^f^rllin COKs c. C.K .e.ions in

ruled h^ein refers .= alX three heavy chain CBRs. or
aXX tLee li.ht chain CDKs or hoth all heavy and all
light Chain CDRs, if approprxate^

CDRS provide the majorrty of contact resrdu

- — — rrtri:: rernredror-

Tor-Lro^tairheavy and li.ht -in se^ence.

and inci.de ^ rarrorrtriire^reTnt::::'

::r;:re::ri:;rnLor ne.trali.in. a^iUty as the
donor antibody fro. which they were

By "Sharing the antigen binding spec f.c.ty or
neutralizing ability" is :»eant, for --^^^ ^^^^
although BC2 ^y be characterized ^ ^ ^
of self-limiting neutralizing act.vrty, a CDR encode
I nucleic acid sequence of BC2 in an appropriate

tructural environment ^y have ^fj,
activity. It is expected that CDRs of BC2 xn such
en'rolents will nevertheless recognize the same
epitope (s) as BC2 ^^^^ ^.^^^^

A "functional fragment xs a pa

• a;:,TnP antiqen binding

region) which retaxns the same -^^^S

specificity and/or neutralxzxng abxlxty
from which the fragment was derxved. ..^^ ^y

An "analog" is an amino acid se<iuence modxfxed by
at le"t one alino acid, wherein said modificatxon can
: Chemical or a substitution or a -arrangement of a
few amino acids (i.e., no more than 10,, whxch
modification permits the amino acid ----- -

the biological characteristics, e.g., antxgen

12

P50438-1

^ affinitv, of the unmodified

soecificity and hxgh attinity/

sequence. Exemplary analogs include silent »utat.ons
which can be constructed, via substitutions, to create
certain endonuclease restriction sites within or
surrounding CDR-encoding regions.

analogs may also arise as allelic variations . An
..allelic variation or modification" is an alteration m
the nucleic acid sequence encoding the amino acid or
peptide sequences of the invention. Such variations or
Ldifications may be due to degeneracy in the genetic
code or may be deliberately engineered to provide
desired characteristics. These variations or
modifications may or may not result in alterations in
any encoded amino acid sequence.

The term -effector agents" refers to non-protem
carrier molecules to which the altered .nti.boa.es
and/or natural or synthetic light or heavy ^^^ams of the
donor antibody or other fragments of the donor antibody
^y be associated by conventional means. Such non-
pritein carriers can include conventional carriers used
m the diagnostic field, e.fl-., polystyrene or other
plastic beads, polysaccharides, e.g., as used m the
BIAcore (Pharmacia) system, or other non-protein
substances useful in the medical field and safe for
administration to humans and animals. Other
agents may include a macrocycle, for chelating a heavy
metal atom or radioisotopes. Such effector agents may
Tso be useful to increase the half-life of the altered
antibodies, e.ff., polyethylene glycol.

For use in constructing the antibodies, altered
antibodies and fragments of this invention, a non-human
species such as bovine, ovine, monkey, chicken, rodent
,e g murine and rat) may be employed to generate a
desirable ir^unoglobulin upon presentment with a human
coagulation factor . preferably factor IX/IXa. X/Xa.
Xl/Xla. Vlll/Vllla, v/va, Vll/VIIa or

13

P50438-1

throntoin/prothrorrtoln or a peptide epitope therefrom,
conventional hybridoma techniques are employed to
provide a hybridoma cell line secreting a non-human mAb
to the respective coagulation factor. Such hybridomas
are then screened for binding using Factor IX/IXa, X/Xa,
Xl/Xia, Vlll/VIIIa, V/Va, Vll/VIIa or
thrombin/prothrombin coated to 96-«ell plates, as
described in the Examples section, or alternatively ».th
biotinylated Factor IX/IXa, x/xa, Xl/XIa, Vlll/VIIIa,
V/Va Vll/VIIa or thrombin/prothrombin bound to a
streptavidin-coated plate. Alternatively, fully huznan
mAbs can be generated by techniques known to those
skilled in the art and used in this invention.

one exemplary, self-limiting neutralising mAb of
this invention is mAb BC2, a murine antibody which can
be used for the development of a chimeric or humanized
molecule. The BC2 mAb is characterized by a self-
limiting inhibitory activity on clotting time. As
measured by the aPTT assay, the effect of the BC2 mAb on
clot time exhibits a maximal value of about 100 seconds.
The BC2 mAb also binds Factor IXa, inhibits Factor IX to
IXa conversion and inhibits Factor IXa activity.
Divalent metal cofactors are required for activity, with
the mAb exhibiting a greater preference for Ca over
Mn=- The observed IC„ in the aPTT assay is
approximately 50 nM. The BC2 mAb exhibits a species
cross-reactivity with rat and is of isotype IgG2a.

Other desirable donor antibodies are the murine
„^s, BCl, 9E4(2)F4 and 11G4(1)B9. These mAbs are
characterized by a self -limiting inhibitory activity on
clotting time. As measured by the aPTT assay, the
effect of these mAbs on clot time exhibits a maximal
value of about 90 to 100 seconds for 9E4(2)F4 and about
80 seconds for 11G4(1)B9. The BCl mAb also binds Factor
ixa, inhibits Factor IXa activity but does not inhibit
Factor IX to IXa conversion. A metal cof actor is not

14

P50438-1

re<iuired for its activity. The observed IC,. £or BCl in
l7:Zr assay is approximately 3B nM. The BCl ^ rs

" "refirher desirable donor antibody characterized
by a self-limiting inhibitory activity on clotting txn,e
Z the .urine ^ H.X.C. As measured by the aPTT assay,
the effect of the HFXLC n,Ab on clot time exhxbxts a
:::ill value of about BO to SO seconds. ^
binds Factor X light chain, and inhrb.ts Factor X/Xa
activity. The observed IC,„ in the aPT^ assay rs
approximately 20 nM. ^-i

vet another desirable donor antibody characterized
by a self-limiting inhibitory activity on clott.ng trme

the murine m^, HFXX . As measured by -J
the effect of the HFXI mab on clot time exhibits a
:::i:ri vame of about lOO seconds. The H.X.C m^ binds
Factor XI and inhibits Factor xl/XIa ^.
Observed IC„ in the aPTT assay is ^^^^^^^
V«lile not intending to be bound to any particular
theory regarding the mechanism of action, these mAbs
appe- to regulate coagulation by a non-competitive or
Illosteric mechanism whereby only partial inhibition is

^^'":;:ts mvention is not limited to the use of the
BCl, BC2, 9E4(2)F4, 11G4(1)B9, HFXLC, HFXI or their

■ v,i^ (n ^ CDR) sequences. Any other
hvpervariable {i.e., ^ut^i = si • ^ -k,,

a^ropriate high-affinity antibodies characterized by a
se!f-Umiting neutralizing activity and corresponding

OHS may be substituted therefor. Xdentification of he
donor antibody in the following description as BCl, BC2,
9E4(2)F4, 11G4(1)B9, HFXLC or HFXI is made for
illustration and simplicity of description only.

The present invention also includes the use of Fab
fragments or F(abM2 fragments derived from mAbs
directed against the appropriate human coagulation
factor or cof actor. These fragments are useful as

15

P50438-1

agents having self -limiting neutralizing actxvxty
against coagulation factors, preferably against Factor
IX/lXa, x/xa, Xi/Xia, Vlll/VIIIa, V/Va, Vll/VIIa or
thrombin/prothronOoin. A Fab fragment contains the
entire light chain and amino terminal portion of the
heavy chain. An F(ab')2 fragment is the fragment formed
by two Fab fragments bound by disulfide bonds. The mAbs
BCl BC2, 9E4(2)F4, 11G4(1)B9, HFXLC and HFXI and other
similar high affinity antibodies, provide sources _ of Fab
fragments and F(ab')2 fragments which can be obtained by
conventional means, e.g., cleavage of the mAb with the
appropriate proteolytic enzymes, papain and/or peps.n,
or by recombinant methods. These Fab and F(ab')2
fragments are useful themselves as therapeutic,
prophylactic or diagnostic agents, and as donors of
sequences including the variable regions and CDR
sequences useful in the formation of recombinant or
humanized antibodies as described herein.

The Fab and F(ab')2 fragments can be constructed
via a combinatorial phage library (see, e.g.. Winter et
al , Ann Rev Irmunol, 12,433-455 (1994)) or via
immunoglobulin chain shuffling (see, e.g., Marks et al.,
Bio/Technology, 10. 779-783 (1992), which are both
hereby incorporated by reference in their entirety,
wherein the Fd or Vh immunoglobulin from a selected
antibody (e.g., BC2) is allowed to associate with a
repertoire of light chain immunoglobulins, Vl (or Vk) , to
form novel Fabs . Conversely, the light chain
immunoglobulin from a selected antibody may be allowed
to associate with a repertoire of heavy chain
immunoglobulins, vh (or Fd) , to form novel Fabs. Self-
limiting neutralizing Factor IX Fabs can be obtained by
allowing the Fd of mAb BC2 to associate with a
repertoire of light chain immunoglobulins. Hence, one
is able to recover neutralizing Fabs with unique

16

P50438-1

3e<^ences (nucleotiae ana a:.ino acia, fro. *e chain

"""rrBro:-o..e. .n...o...s ..s.r.... a.ove .a.

The mAD variable heavy ana/or

contribute ..^^worlc se^ences, CDR

rerrrroL-^ente, a—

-^^''^ "tro;t:rni::rrro:rartra li^oaie. .hi.h
rrhrac:::i:er:. l ^^..^^ --in. .eci^icit. o.

the donor antibody. invention, or

nucleic acia se<^ences encoding the cnHso ^^^^^^^^

"^^rLc litate insertion of ^uta.eni.ea COK ana/or
which tacixitdLc -rf^aions can be

- - — - i": >
::n "s: :fr;r.:rrr. ... >.

SEQ ID NOs: 8, 9 ana 10 ^^.^.es of the BC2

... nucleic ana ^.no ac a e^^e^ ^ ^

-^^^^^^^^^ -o. thi. re.ion are listea

code various coding sequences may be constru

Toae the variable heavy ^^^^ Til a.

sequences and CDR sequences of the xnve

17

P50438-1

orCDRs can be .seS to produce altered a„txbod.es. e.^..
cLLric or humanized antibodies or other engxneered
TZo^ies of this invention when operatively conO^rned
with a second immunoglobulin partner.

Xt Should be noted that in addition to --"^
.ucleic acid sequences encoding portions of the altered
antibody and antibodies described here.n, °^^er such
nucleic acid se<^ences are encompassed by the present
invention, such as those complementary to ^^"^
CDR-encoding sequences or complementary to the modrf.ed
hLn frameworlc regions surrounding the CDR-encod.ng

useful DNA sequences include those sequences
rch lybr!:::: under stringent hybridization conditions

Cf^e T. Maniatis et al . ,

ho t-He DNA sequences. bee,

!:o lecxar Cloning: . Laboratory Manual-, Cold Spr.ng
Harbor Laboratory (1982,. PP- 387-389. An exa:^le
one such stringent hybridization condition rs
hybridization at 4XSSC at 65"C. followed by a washing in
O.IXSSC at eS'C for one hour. ^1'-""^;^^^-. ^^^^^
exemplary stringent hybridization condition is 50%

4XSSC at 42°C. preferably, these hybridizing
formamide, 4XSSC ac v.. ^. -.^^

sequences are at least about 18 nucleotides m

DNA S

--ll^irld-i— Lrirmrieri::-can encode altered
...ibodies Which include engineered anti^dies such a.
Chimeric antibodies and humanized antibodies^ . ^^s.r^
altered i,»unoglobulin coding region -""-^^"'^
encoding regions that encode peptides having the antigen
specificity of a Factor IX/IXa, X/Xa, KX/XIa,
Viri/VIXIa, V/Va, VIX/VIIa or thro:rtoin/prothrombin
antibody, preferably a high affinity ant body sue, a
provided by the present invention, inserted into
18

P50438-1

-^^v ^uch as a human framework or
immunoglobulin partner sucn

human immunoglobulin variable regxon^

preferably, the first immunoglobulin partner

operatively linked to a second inonunoglobulin partner^

of interest, for example an Fc region. Second
iLno.lo.uiin partners .ay also include se„
encoding anotl^er immunoglobulin to "''^^ ^^^^f
heavy chain constant region is fused xn frame or by
TZ of a linker sequence. Engineered antxbod.es
Tircted against functional fragments -alog^ of
coagulation factors may be designed to elrcxt enhanced
binding with the same antibody.

The second immunoglobulin partner may also be
associated with effector agents as defined above
trcluding non-protein carrier molecules, " f/^^^
second i^unoglobulin partner may be operatrvely Ixnked

bv conventional means. ^^loVMilin
Pusion or linkage between the second xmmunoglobulxn
partners, e.g., antibody se<^ences, and the effector
agent may be by any suitable means, e.g., by
rrentional covalent or ionic bonds, protern fusxons,
or hetero-bifunctional cross-linkers, e.g.,
carbodiimide, glutaraldehyde and the like. Such
technigues are known in the art and are described rn
conventional chemistry and biochemistry texts^

Additionally, conventional linker -^^^^ fj;*^^
simply provide for a desired amount of space between the
simpj-y effector agent may

second iimunoglobulin partner and ^^^^''""^ ^

— r":::re:-^^^^^^^^^^

- "^1^. id TeUces for the molecules of
the invention may be modified by techniques known to
those skilled in the art to enhance expression.

19

P50438-1

A preferred altered antibody contains a variable
heavy and/or light chain peptide or protein sec^ence
having the antigen specificity of BC2, e.g., the Vh

aid V. Chains. Still another desirable altered ant.body
of this invention is characterized by the amino acrd
sequence containing at least one. and preferably all of
the CDRs of the variable region of the heavy and/or
light chains of the murine antibody molecule BC2 w.th
the rer,>aining secjuences being derived from a human
source, or a functional fragment or analog thereof.

in a further embodiment, the altered antibody of
the invention may have attached to it an additional
agent For example, recombinant DNA technology may be
used to produce an altered antibody of the invention m
which the FC fragment or CH2 CH3 domain of a complete
antibody molecule has been replaced by an enzyme or
other detectable molecule (i.e.. a polypeptide effector
or reporter molecule) .

The second immunoglobulin partner may also be
operatively linked to a non- immunoglobulin peptide,
protein or fragment thereof heterologous to the CDR-
containing sequence having antigen specificity to a
coagulation factor, preferably to Factor IX/IXa, X/Xa,
Xl/Xia, Vlll/VIIIa, V/Va, Vll/VIIa or

thror^in/prothrombin. The resulting protein ^^"^^^^^
both antigen specificity and characteristics of the non-
i^nunoglobulin upon expression. That fusion P-"-'
characteristic may be, e.g., a functional characteristic
such as another binding or receptor domain or a
therapeutic characteristic if the fusion partner is
itself a therapeutic protein or additional antigenic
characteristics.

Another desirable protein of this invention may
comprise a complete antibody molecule, having full
length heavy and light chains or any discrete fragment
thereof, such as the Fab or F{ab')2 fragments, a heavy

20

P50438-1

chain dimer or any .ninimal recorrtoinant fragments thereof
chain ^i^gi^.chain antibody (SCA) or any

such as an Fv or a siiiy-^ cc.l*=rted
other raolecule with the same specifxcrty ^^'^^ l^^^^'^^
donor r^. e.,., ^ BCl, BC2, 9K4,2,.4 l^-^^'^*-/;^
HFXLC or HFXI. such protein may be used in the form
al altered antibody or may be used in its unf used form.

vmenever the second immunoglobulin partner rs
derived from an antibody different from the ^onor
Ttibody, e.... any isotype or class of i^nunoglobulin
™r. or constant regions, an engineered antrbody
results. Engineered antibodies can comprise
i^aunoglobulin ,Ig, constant regions and -^-^^^

=-r;r.rjr

donor antibody, e.g., the anti-Factor ix/ixa, X/xa,
Xl/Xia, Vlll/VIIIa, V/Va, Vll/VIIa or
.hrombin/prothro^in antibodies described ^
^ ^ f^e^l e-tions , substitutions,
addition, alterations, e.g., deletions,
ridditions, of the acceptor ^ lig^t and.or heavy
variable domain frameworK region at ^-."-1-; -/^J^
amino acid levels, or the donor CDR — -"^^f
in order to retain donor antibody antigen binding

^'^'^"suirengineered antibodies are designed to employ
one (or both, of the variable heavy and/or light chains
of the coagulation factor (optionally modified as

CDRs. The engineered antibodies of the invention
exhibit self-limiting neutralizing activity.

such engineered antibodies may include a humanized
antibody containing the frameworlc regions of a selected

immunoglobulin or subtype or a chimeric antibody
containing the human heavy and light chain constant
regions fused to the coagulation factor antibody
functional fragments. A suitable human (or other
animal, acceptor antibody may be one selected from a

21

PS0438-1

n ^.f^hase eg., the KABATO database, Los
ccnventxonal ^.^.^.e,JJ - ^^^^^^^^ ^ ^^^^^

Alamos database, and Swiss fro

to the nucleotide and amino acid sequences of the donor
antibody A human antibody characterized by a homology
: ::: Irame„or. regions o. the donor antibody <on^n
amino acid basis, may be suitable - P-^J^J
chain variable framework region for .nsertxon of the
lonor CDRs. A suitable acceptor antibody capable of
riing light Chain variable -^^mewor^^^^^^^^^^^^^

required to originate from the same acceptor

preferably, the heterologous framework and constant
regions are selected from human ^^^^^^^^
and isotypes. such as IgG (subtypes 1 through 4) , IgM.

aTxgB. However, the acceptor antibody need not
cimprise only human immunoglobulin protein ---^-^
Por instance, a gene may be constructed xn whxch a DNA
^^Hlna part of a human immunoglobulin chain
Ts^:::: ZTZ iZ^r... encod^g a non-i„Xobulin
lli" acid sequence such as a polypeptide effector or

reporter molecule.

A particularly preferred humanized antibody
contains CDRs of BC2 inserted onto the framework regions
of a selected human antibody se^ence. For neutralizing
iLized antibodies, one. two or P-^^-
from the Factor IX antibody heavy chain and/or light
Chain variable regions are inserted into the framework

regions of the selected human antibody sequence,
replacing the native CDRs of the latter

preferably, in a humanized antibody, ^^^^J^"^^^^
aomains in both human heavy and light chains have been
enaineered by one or more CDR replacements. It
Tosslble to use all six CDRs, or various co.a.inations of
les than the six CDRs. Preferably all six CDRs are
replaced. It is possible to replace the CDRs only in

22

P50438-1

the wan heavy chain, using as light chaxn the
^^odified light Chain frc. the huraan acceptor antibody.
Still alternatively, a compatible light cha.n :«ay be

cted fro» another human antibody by ----
conventional antibody databases. The ^- "^^^'^^

engineered antibody may be derived from any suitable
acceptor human immunoglobulin.

The engineered humanized antibody thus preferably
has the structure of a natural human antibody or a
fragment thereof, and possesses the combination of
pro^rties retired for effective therapeutic use^e. ^. .
treatment of thrombotic and embolic diseases in man

Most preferably, the humanized -tibodies have a
heavy chain amino acid se^ence as set forth in SEQ ID
To .1. 52, or 89. Also most preferred are humanized
antibodies having a light chain amino acid se^ence as
set forth in SEQ ID NO: 44, 57, 62, 74 or 99.

particularly preferred is the ^---^^

249413 v,here the heavy chain has the ammo -"^-^J"
as set forth in SEQ ID m-. 31 and the light chain has
tL alino acid se.^ence as set forth in SBQ ID NO: 44.
tl!o r-articularly preferred is the humanized antibody SB
r4inrere t JhLvy Chain has the amino -id se^ence
as set forth in SEQ ID NO: 52 and the light chain ha
the amino acid se^ence as set forth in ^ J-^^^'
Mso particularly preferred is the humanized antibody SB

- :::: t:^

— tit:— ^^^^^
-rr::rn%r7D:-::d-^^^^^^^^^^

L amino acid sec^ence as set forth in ^^^^
Mso particularly preferred is the J^";;^^//,
257731 where the heavy chain has the ammo
as set forth in SEQ ID NO: 52 and the light cham has

23

P50438-1

..e an,ino acid sequence as set fort^ ^^^^^f ^J^'^^
Also particularly preferred is the humanized --^^^'^^
' ,32 Where the heavy chain has the a»ino -^^--J"
as set forth in SEQ ID NO: 89 and the light charn has
:L a^ino acid sequence as set forth in SKC .0:

It will be understood by those skilled -"^he art
that an engineered antibody ™ay be further modrfred by
Changes in variable domain a»ino acids without
necessarily affecting the specificxty and h gh aff.nrty
of the donor antibody (i.e., an analog). xs
aLicipated that heavy and light chain amino acxds :»ay
be substituted by other amino acids erther m the
variable domain frameworks or CDRs or both. These
substitutions could be supplied by the donor antxbody
consensus se<^ences from a particular

in addition, the constant region may be altered
enhance or decrease selective properties °^
molecules of this invention. For example,
binding to .c receptors, or the ability to bxnd nd
activate complement (see, e.g., imgal et -i'' "°
x™.unoI, 30, 105-108 ,I9«, , Xu et

269 3469-3474 (1994), Winter et al., EP 307434-B).

^ altered antibody which is a chimeric antxbody
differs from the humanized antibodies described above by
providing the entire non-human donor antibody heavy
chain and light chain variable regions, xncludxng
framework regions, in association with human
i™„unoglobulin constant regions for both chaxns^ It
anticipated that chimeric antibodies whxch retaxn
additional non-huxnan se^ence relative to
antibodies of this invention may elicit a sxgnxfxcant
immune response in humans.

such antibodies are useful in the preventxon and
treatment of thrombotic and errbolic disorders, as
discussed below.

24

P50438-1

preferably, the variable light and/or heavy chain
sequences and the CDRs of ^ BC2 or other

e..., BCl. 9B4(2,.4, lia4<l,B9, 3.
and their encoding nucleic acid se-juences. "^^^^^^
in the construction of altered antibodies, preferably
humanized antibodies, of this invention, by the
following process. The same or ^i-^^^^-^^^^^^^^
also be employed to generate other embodiments of th.s

invention. av. o

J A-^r^ a c;^l e^cted donor mAto, e.g.,
A hybridoma producxng a seieccea

the murine antibody BC2, is conventionally cloned and
the DNA of its heavy and light chain

obtained by techni^es known to one of slcrll .n the art,
e <, the techniques described in Saittrook et al.,
..M^llcular Cloning: A Laboratory Manual " , 2nd ed.t.on,
cold spring Harbor Laboratory (1989) . The ---=1^
heavy and light regions of BC2 containing at least the
CPR^ncoding regions and those portions of the acceptor
mAb light and/or heavy variable domain framework regrons
required in order to retain donor mAb bindrng
specificity, as well as the remaining iirmunoglobul.n-
Trived parts of the antibody chain derived from a human
i^unoglobulin, are obtained using polynucleotrde
primers and reverse transcriptase. The CDR-encod.ng
regions are identified using a Icnown database and by
comparison to other antibodies.

A mouse/human chimeric antibody may then be
prepared and assayed for binding ability. Such a
chimeric antibody contains the entire non-human donor
antibody V„ and Vl regions, in association with human Ig
constant regions for both chains.

Homologous framework regions of a heavy chain
variable region from a human antibody are ^^^^^^^^
using computerized databases, e.^., KABA^ and a human
antibody having homology to BC2 is selected as the
acceptor antibody. The sequences of synthetic heavy

25

P50438-1

Chain variable regions containing the BC2 CDR-encoa.ng
regions within the human antibody frameworks are
designed with optional nucleotide replacements rn the
framework regions to incorporate restriction sxtes
This designed sec^ence is then synthesized usrng long
synthetic oligomers. Alternatively, the designed
sequence can be synthesized by overlapping
rifgonucleotides, amplified by P^-erase chain reaction
(PCR). and corrected for errors. A suitable light chain
variable framework region can be designed in a similar

manner- ,

A humanized antibody may be derived from the
Chimeric antibody, or preferably, made ^vnthetically by
inserting the donor mAb CDR-encoding regions from the
heavy and light chains appropriately within the selected
heavy and light chain framework. Alternatively, a
humanized antibody of the invention may be prepared
using standard mutagenesis techniques. Thus, the
resulting humanized antibody contains human framework
r::ions and donor mAb CDK-encoding regions. .here may
be subsequent manipulation of framework residues,
resulting humanized antibody oan be expressed in
reco^cO^inant host cells, e.g., COS, CHO or myeloma cells .
Other humanized antibodies may be ^^-^-"^^l^.^^l^Hl^
technique on other suitable Factor IX-specific or other
coagulation factor-specific, self-limiting,
neutralizing, high affinity, non-human antibodies.

A conventional expression vector or recombinant
plasmid is produced by placing these -^^"^ --"^"^f^^
for the altered antibody in operative association with
conventional regulatory control sequences "P^^^ f
controlling the replication and expression m. and/or
secretion from, a host cell. Regulatory sequences
include promoter se,^ences, e.g., CMV promoter and
signal sequences, which can be derived from other known
Ttibodies. Similarly, a second expression vector can

26

P50438-1

,e proaucea .avin. a DNA se^ence »l.ic. encode, a
..„p.e.en.a.. ^^^^^^ ^^^^^ ^^^^^^"^

conventional technic^es w.th both the £

vectors ,or si»ply ---ff^^f , L invention

create the transfected host ^^..^^i^ Ught and

rrr,:: ":r:-. " "

other altered antibodies and molecules of th.s

'"^'"surtable vectors £or the cloning and subclonin.
steps^ploved i. the .ethods and oonstruct.o h
compositions Of this Of

fro. .....

r—i; rrrpi:::!. readH. ha. a„ ^ndance Of
not a limiting factor in this invention.

27

P50438-1

Similarly, the vectors eT^loysd for expression o£
eUeere. a.ti.oaies oc r. t..^r^n..o„

- cfor" tIc Te rs also contain selected

conventronal vector. The ^^^.^.^^ „hich

regulatory sequences (such ^J^J j^^.^.^.o^ous

airect the "-^^-"""^"/JJ/:" s. These vectors
se^ences in selected

efected i:™»unoglo.ulin se^ences -o^^^JJ
insertion of desirable restriction srtes for ready

manipulation. characterized by

The expression vectors may also D
genes suitable for amplifying expression of the
^ ^ rr the maminalian

l^eterologous DNA other preferable

aihydrofolate -<^---;;";„ °T 3e.uence, such

vector sequences rnclude a poly A g

.s from bovine growth sectors
nromoter sequence (betaglopro) . The exp
Tslful here" may be synthesized by techniques well
known to those skilled in this art.

The components of such vectors, e.g.. replico ,

— — ^^^^^^^^^

r synthLi^ed by .nown p—

airecting the expression and/or secretron of the P
„f the recombinant DNA in a selected host. Other

Z expression vectors of which numerous types
appropriate expression K-cterial, insect,

are known in the art for —^^^"'^f
yeast and fungal expression may also be selecte

TeTesent invention also encompasses a cell line

^sfectL with

28

P50438-1

.He Cloning and other manipulations of ^-^ J^-f/^^
vectors are also conventional. However,

• ^ r^-f TT coll are usea tor
rells from various strains of E. co±i a
replication o. t.e cloning vectors an. other steps .n

the construction of altered antibodxes of thrs

invention.

suitable host cells or cell Imes for the
expression of the engineered antibody or altered

of the invention are ---- ^ ; ^--^^^
such as CHO, COS, a fibroblast cell (e.g., 3T3) an
ireloid cells, and .ore preferably a CHO or a .yelo.d

Hu»a„ cells .ay be used, thus enabling tbe
molecule to be modified with human -Ivcosylatxon
patterns. Alternatively, other eukaryotxc cell Ixnes

be employed. The selection of
host cells and methods for transf ormatxon, culture
rUfication, screening ^ -rr"s::^^oo.
purification are known in the art. See, e.g ,

et al. , supra.

Bacterial cells may prove useful as host cells
.or the expre s. the reco—^^

of proteins expressed in bacterial cells to be xn an
unfolded or improperly folded form or in a non-
glycosylated form, any recoKtoinant Fab produced .n a
fafterLl cell would have to be screened for retentron
of antigen binding ability. If the molecule expressed
t the bacterial cell was produced in a P-P"^- "^^^^

.Lm, that bacterial cell would be a ^-"^^^^

Por example, various strains of E. CO a

expression are well-lcnown as host cells .n the fxeld

biotechnology. Various strains of B.

Streptomyces, other bacilli and the Ixke may also be

employed.

29

P50438-1

«>ere desired, strains of yeast cells )cnown to
..ose skilled in t.e art are also available as .o^t
cells, as well as insect cells, e... Prosoph. a and
Leoidoptera and viral expression systems. See, eg.
Lepiaopcexo 277-298, Pleniim

Miller et al.. Genetic Engineering, 8, 277

required to produce the host cells ^
culture methods necessary to produce the -"--^
antibody of the invention from such host cells are all
conventional techni^es . Likewise, once -^<'^-'^,^,'J''^^^
altered antibodies of the invention may be purrfred from
the cell culture contents according -^^-"^"^
procedures of the art, including ammonium sulfate
preciSi-^i-. -«^-^^ chromatography.
Tel electrophoresis and the liKe. Such techniques are
Within the skill of the art and do not limit this

'"^nlranother method of expression of the humanized
antibodies may utilise egression — 3^,.

animal, such as described m U. S. Patent
This relates to an expression system using the animal s
casein promoter which when transgenically --J-
into a maiM^al permits the female to produce the desired
recombinant protein in its milk.

once expressed by the desired method, the
engineered antibody is then examined for in vitro
activity by use of an appropriate assay. Presently,
conreLIonal K.XSA assay formats are employed to assess
g:alitative and .^antitative binding of the engineered
antibody to Factor IX or to other -P^-^"^^^
coagulation factors. Additionally,

assays may also be used to verify neutralising efficacy

30

P50438-1

evaluate the persistence of the engineered antxbody rn
the body despite the usual clearance mechanxsms.

Pollowin. the procedures described for hu^anrzed
antibodies prepared from BC2 , one of skill in the art
Tay all construct humanized antibodies fro„ —
Iltibodies, variable region sec^ences and CDR peptrdes
irs ribed herein. Engineered antibodies can be produced
„ith variable region frameworks potentially recognrzed
as -self by recipients of the engineered antibody
Tinor modifications to the variable region ;
can be implemented to effect large increases « -'J^
binding without appreciable increased immunogen.city
the recipient. Such engineered antibodies may
effectively treat a human for coagulation factor-
mediated conditions. Such antibodies may also be useful
in the diagnosis of such conditxons.

This invention also relates to a method for
inhibiting thro.4.osis in an animal, particularly a
:Lan, Which comprises administering - effectrve dos
of an anti-coagulation factor monoclonal antxbody havxng
self-limiting neutralizing activity in co:*xnatxon wxth
a plasminogen activator. Co:<^ination therapy enhances
thrombolysis at sub-optimal

plasminogen activator decreasing the txme to
of blood flow and increasing the frequency and total
duration of vessel reperfusion. In contrast to heparxn.
cox,*ination therapy does not significantly perturb
normal hemostatic functions and spares fxbrxnogen,
plasminogen and alpha-2-antiplasmin levels.
Accordingly, this invention also relates to a method
reducing a re,^ired dose of a throTtoolytxc agent xn
treatment of thrombosis in an animal comprxsxng
adllnistering an anticoagulant specifically targetxng a
component of the intrinsic coagulation pathway xn
combination with the thrombolytic agent.

31

P50438-1

preferably, the coagulation factor is from the
intrinsic or co,«aon coagulation pathway. Most
preferably, the anti-coagulation factor ■nonoclonal
pretera y, anti-Factor Ixa, anti-

antibody IS an anti-Factor i p.^tor
Factor X, anti-Factor xa, anti-Factor XI, anti Factor
X a anti-Factor VXXX, anti-Factor VXXIa, ----^J:

Factor VII, anti-Factor Vila, anti
rhlll:: or anti-prothro^^m. The ^ can include one
or more of the engineered antibodies or altered
antibodies described herein or fragments thereof.
Preferably, the plasminogen activator is tPA,

re:to.inase, urokinase or tPA variants as described

— ---^-^r:r°";/rcrti r.s^ -B^r;

14585 (1997); FUDlse et al . , Circuia

,,997,, coombs et al., .Biol C.e™, ^"'/f ^^^^^^
,1998); van de Werf at al., A. Heart J, 137, 786
,1999) . Particularly preferred is tPA.

Alternatively, acetylsalicylic acid can be
administered in con*ination with the anti-coagulation
:rtor monoclonal antibody. Xn some cases, combine ion
therapy lowers the therapeutically effective dose of
anti-coagulation factor monoclonal antibody.

L therapeutic response induced by the use of
the molecules of this invention is produced by the
binding to the respective coagulation factor t^^
s^blc^ent self-limiting inhibition of the coagulation
cascade Thus, the molecules of the present invention,
:Zln preparations and formulations appropriate for
Therapeutic use, are highly desirable for person3
susceptible to or experiencing abnormal clotting
acti'ty associated with, but not limited to, myocardial
infarction, unstable angina, atrial fibrillation,
strolce, renal damage, pulmonary e^olism, deep vein
:hromb;sis and artificial organ and prosthetic implants.
A particularly preferred use is in myocardial
infarction.

32

P50438-1

The altered antibodies, antibodies and fragments
thereof of this invention may also be used ^n
on lotion with other antibodies, particularly h^n
reactive with other markers (epitopes, responsible
Tr the condition against which the engineered ant.body
of the invention is directed.

The therapeutic agents of this invention are
believed to be desirable for treatment of abnormal
Clotting conditions from about 1 day to about 3 weeks,
or as needed. This represents a considerable advance
over the currently used anticoagulants ^^^^^^
warfarin. The dose and duration of treatment relate,
the relative duration of the molecules of the present
inveltion in the human circulation, and can be adjusted

Tone o* ^-^'^ ^^'^ ''^^^"^"'"^ 11

being treated and the general health of the patxent.

The mode of administration of the therapeutic
agents of the invention may be any suitable --^e whrch
Tllvers the agent to the host. The altered antrbod.es,
antibodies, engineered antibodies, and fragments
thereof, plasminogen activator and pharmaceutical
colposi ions of the invention are particularly useful
for parenteral administration, i.e., subcutaneously .
intramuscularly, intravenously or intranasally^

Therapeutic agents of the invention may be prepared
as pharmaceutical compositions containing an
lount of the engineered .e.g.. humanized, antibody of
the invention and plasminogen activator as active

ngredients in a pharmaceutically acceptable carrier^
Alternatively, the pharmaceutical compositions of the
invention could also contain acetylsalicylic acid,
the prophylactic agent of the invention, an acraeous
suspension or solution containing ^^^^^-^"^ ,^ ^
antibody, preferably buffered at physiological pH, in
orl ready for injection is preferred. The compositions
for parenteral administration will co,»only comprise a

33

P50438-1

r^vc^f f:>rablv an aqueous carrier,
acceptable carrxer, preferably ^ ^ o.4%

variety of aqueous carriers may be employed eg ,
variety o^ These solutions are

saline. O'^* ^^-^-^/"tro.'L.iculate -natter. T.ese

sterile and ^^^^^/^^f^/J J,,„.entional, well Kno„„
solutions may be sterilizea oy

.;„,.i.„i-.

or 20* by -^^*'\-^^7;,^^;^::'^'^L.oLing to the
fluid volumes, viscosities, etc.,

particular r.ode of -^'^'^';"''°\Z'Z l,\^^ invention
^us, a Pharmaceutical ^^'^^^^^^'^jl^^ ,o„,,iu

rLTtiretLreH^^^^^^^^ ^ - -

rut^rorm. e... a^ut --^rl^rred
preferably, about 5 m. to about 25

.„.,.ody Of - rtion'f:: intrlvenous infusion

composition of tne mve sterile
..uld be made up to .SO ^o. .

rfrra^^r" - » — : -^^'^-^^

preter ^ n^ethods for preparing

of the invention. Accua are well known

parenterally administrable -----^ "^^^

- wnl -.--ent to t, -1 led^^

rr^relLrrcire^^ ed., Mac. Publisbin.

--Tt\r:re:errertrre-t.erapeutic a.ent. of tbe
invention, w.en in a pbarmaceutical preparation, be

P50438-1

... .... .= """"r .... .. I p....-

:r.;"i:*r.„„ .... - — • -

employed .

The present Invention will now be described with
.e^erlnce to the following specific. non-l.n>.trna
examples .

35

P50438-1

..a scr g of >n.i-F...or TX Monoclonal .

ani-i bodies

Female Balb/C mice were injected with human factor
IX purified as described in Jenny, R. et al.. Prep
Bloche., 16. 227-245 (1986). Typically, each mouse
received an initial injection of 100 ug P"^""
dissolved in 0.15 m. phosphate-buffered salrne ,PBS, and
mixed with 0.15 m. complete Preund.s ^^^^^^^^
i^unizations of 50 ug protein in 0.15 mL PES w.th 0 15
m. incomplete Preund-s adjuvant were given approximately
biweelcly over a 2-3 month period. After the frnal
boost, the mouse received 50 ug of Factor IX rn PBS
three days before spleen/myeloma cell fusxons. Spleen
cells were isolated from an immunized mouse and fused
with NS-1 myeloma cells (Kohler, G. e£ al . , Bur J
Immunol, 6. 292-295 (1976,) using polyethylene glycol as
described by Oi, V.T. et al . in -Selected Methods .n
cellular Immunology," Mishell, B.B. and Shigix, S.M..
eds . Freeman Press. San Francisco. Following the
fusion, the cells were resuspended in RPMI 1640 med.a
containing 10% fetal calf sera and alicjuots were placed
in each well of four 24-well plates contarnrng 0.5 mL
peritoneal lavage cell-conditioned media. On the ^
following day, each well received 1.0 m. o 2 x 10
hypoxanthine. 8 x lO" M aminopterin and 3.2 x 10 M
t^idine in RPMI 1640 media containing 10% fetal calf
se^ The cells were fed every 3-4 days by removing
half of the media and replacing it with fresh medra
containing 1 x IQ- M hypoxanthine and 1 . 6 x 10 M

thymidine. 1 0 mL of hybridoma

Approximately two weeks later, i.u mu
medium was removed from each well and tested for antr-
Factor IX antibodies using an E.ISA assay as descrxbed
by Jenny, R.J. at al . in Keti, Bmy^ol. 222. 400-416
,1993). Briefly, factor IX was immmobilized onto

36

P50438-1

11. of 96-well microtiter plates. Hybrxdoma
plastxc wells of 96 wel antibody were then

supernatants or dilutions of purxfxed

substrate o-dianisidine^ ,„,ibodies were

wells containing anti-Factor
subcloned by limiting dilution and grown in 96 well
nlates supernatant from the cloned hybridoma cell
plates, bup antibody to Factor IX by the

^„lt-iires were screened for antmouj

Hssay described above and cells ^^^J^

.y^ridomas were — -^iZrul^s^ re
nitrogen and then grown as ascitic

, ,^ , ;^..-rn, . rna.ion Facto r

The effect of increasing concentrations of anti-
coagulation factor antibodies on activated partial
thro^oplastin time (aPTT, of human plasma was
aetermined in a fibrometer (Becton-DicHinson
microbiology Systems, '^-^^^rT^Z:;^
Baxter reference procedure LIB0293 J, i/^
(Baxter Scientific, Edison, New Jersey)

prior to the start of the experiment, 2 to 3 ml.

. tube were placed into the heating

n no M raCI in a 5 mL tuPe weie fi

Of ;he ..rometer^ ^an Plasma sa^^^^^^^^^

reference Plasma (^^erican Diagnostics, Greenwich,

^"""^rfirctionated heparin from porcine intestinal
T ci- Tnnis Missouri), low

:re::irrigrh:::i:f.rp:r;ine intestmai mucosa

Cenox*, enoxaparin sodium. Rhone-Poulenc Korer

P50438-1

t-^rals Collegeville, Pennsylvania) or mAb
Pharmaceuticals, Coixey« .^^t-elv BO uM

anticoagulant was included as a ^^^^l^^^ ^.^^ ,oO

TWO fibroTube* fibrometer cups were fiHea w
.1 tert Plasma or 100 ul test plasma with anticoagulant
L 125 ul of actin activated cephaloplastin reagent
r^lirreagent. fro. rabbit brain ^---^
aoid, available from Baxter Screntrf xc) , ^respectw y
and placed in the fibrometer wells at SVC.

. 100 ul of actin reagent was

after one minute, luu u± oi. a
..ansreLd to a plasma-containing cup and tb contents
.ixed several times with a pipette. After
incubation, 100 ul of CaCl., prewarmed at 3^ C
.o the plasma-actin reagent mixture using ^ -toma c
Pipette/Timer-trigger <Becton-DicHinson. .he ^^^^^^

rcLrtinr:: r-tir orfi: i con e nations

^-o^al assay volvune of 30U ux.
of anticoagulant m the "tal as y ^^^^^ ,^

The nominal concentration of Factor ix

"-^T^^-results shown in .ig. 1 -"-"--f/jr
increasing concentrations of mu-e an i-.act^^

XX n^s BCl and BC2 on ^3
inhibit clotting by prolonging the aPTT and b

:each a final ^ -/^-aL^-sriTo: BCran^^C^ ,

rp::reir:r t:: drff:ence m the maximum response

- - -^--rBrl Lrret^hrarTTabout to

=r::rr ol Ther hand, increases the aPTT by

; s-fold to about 90 sec. The therapeutic target zone
.sed in anticoagulant therapy with -P^"; ^
highlighted. The results indicate that the
bracket the heparin therapeutic aPTT range.

P50438-1

The properties of n^s BCl and BC2 are

■ ^ in Table I Each of the BC mAbs recognizes
summarxzed xn Table I ^^^.^^
both the zymogen, Factor IX, as well a
protease, Factor IXa, but only BC2 .s -P^^^^;
Locking both zymogen activation as well ^^ J^TZ^
activity. BCl and BC2 were found to -oss-react w
Cynomologous monkey Factor IX. Additionally, BC2 also
cross-reacted with rat Factor IX.

Table I.
IX mAbs

sugary of in vitro Properties of ^nti-Factor

BCl

BC2

Binds Factor IX

Binds Factor IXa

inhibits IX to IXa
conversion

Inhibits IXa activity
in Xase complex

Cofactor requirement

yes
yes

yes

yes
yes
yes

divalent
metals

aPTTmax X 100%
aPTTnormal

ICso,

Species cross-
reactivity

150

-35

Ca > Mn
350

monkey rat, monkey
IgGl IgG2a

39

P50438-1

The results shown in Fig. 2 demonstrate the effect
of increasing concentrations of the anti-Factor IX mAbs
9E4 (2) F4 and 11G4(1)B9 on aPTT clotting times. The
plasma for the assay was diluted to one-half the normal
concentration, giving an initial aPTT of 45 seconds.
Both msbs inhibit clotting by prolonging the aPTT and
both mAbs reach a final saturating effect on the aPTT.
saturating concentrations of 9E4,2)F4 and
increases the aPTT to -90 to 100 seconds for 9E4(2)F4
and to -80 seconds for 11G4(1)E9. The results indicate
that the two mAbs are at the upper end of the heparxn
therapeutic aPTT range.

The results shown in Fig. 3 demonstrate the effect
of increasing concentrations of the anti-Factor X mAbs
HFXLC (vs. light chain epitope). HFXHC (vs. heavy charn
epitope) and the anti-Factor XI mab HFXI on aPTT
clotting times. These :^s were obtained ^^^^
Research Laboratories (South Bend, IN) . The mAbs HFXLC
and HFXI inhibit clotting by prolonging the aPTT and
both mAbs reach a final saturating effect on the aPTT.
The IC value for HFXLC is -40 nM; saturating
concentrations increase the aPTT to -60 seconds. The
IC value for HFXI is -20 nM; saturating concentratxons
increase the aPTT to -100 seconds. The results indicate
that HFXLC is within the heparin therapeutic aPTT range
while HFXI falls at the upper end of the heparxn
therapeutic range. The mAb HFXHC had no effect on aPTT

clotting times.

Self-limiting prolongation of the aPTT was also
observed with antibodies to Factor VIII. the cof actor to
Factor IXa. For example, the anti-human Factor VIII
antibody. SAF8C-IG, purchased from Affinity Biologxcals,
inc increased the aPTT to a maximum of about 65 sec.
Half-maximal prolongation of the aPTT was achieved wrth
about 100 nM antibody.

40

P50438-1

Example 3

' 7n Lder to ev aluate the efficacy of anti-Factor IK
antibodies in prevention of arterial throinbosis, the rat
carotid artery thrombosis model as reported by
™cher et al. in . Car.ic ..ar., 526-533 (1393,

was adapted. This model consists of segmental -3"-^ ^°

the carotid endothelium by oxygen radicals generated by

Feci, solution applied on the surface of the carotrd

artery .

in brief, rats were anesthetized with
pentobarbitone sodium, the jugular vein cannulated for
intravenous injections and the left femoral artery
cannulated for blood pressure and heart rate monrtorxng.
The carotid artery was isolated by aseptic technxque vxa
a surgical incision in the neck and equipped wxth a
magnetic flow probe for blood flow measurement. After a
period of stabilization, baseline parameters were
established for the following variables: carotid blood
flow, arterial pressure, heart rate, activated partial
thromboplastin time (aPTT) and prothrombin trme (PT)^
Thereafter, a premeasured Whatman filter paper soaked xn
50% Feci, solution was placed on the carotid artery for
15 minutes for complete injury of the underlyxng
endothelial cells. After removal of the FeCl. soaked
paper, the experiment was followed to completion over
minutes. At the end of the experiment, the carotxd
throTtous was extracted from the carotid artery and

weighed. .

All agents were administered 15 minutes prxor to
the onset of carotid injury. The following treatments
were examined and compared to the Factor IX mAb BC2 .

1 Heparin: 15, 30, 60 or 120 U/kg bolus,
followed by infusion of 0.5, 1, 2 or 4 U/kg/min,
respectively over 60 minutes

60

41

P50438-1

2. Acetylsalicylic acid (ASA, aspirin): 5 mg/kg

bolus , -

3. Anti-Factor IX mAb BC2 : 1 , 3 or 6 mg/kg bolus,

followed by infusion 0.3, 1, or 2 ug/kg/min,
respectively over 60 minutes

4. Heparin: 30U/kg bolus + lU/kg/min + ASA at 5

Anti-Factor IX mAb BC2 : 1 mg/kg .0.3 ug/kg/min
+ ASA at 5 mg/kg

Figs 4 and 5 demonstrate the comparative
pharmacology of the ant i -coagulant /thror^otic regimens
by showing the effect of heparin, ASA and Factor IX mAb
BC2 on aPTT (Fig. 4) and PT (Fig. 5).

The key index for bleeding diathesis, aPTT, was
used as the primary criterion for evaluation of efficacy
versus bleeding liabilities of the anti-
coagulant/ thrombotic agents used in the study. The
results in Fig. 4 demonstrate the dose -dependent
prolongation of aPTT by heparin with maximal
prolongation of the clotting time, beyond the test
limit, at the two higher doses. ASA alone dxd not
significantly increase aPTT but in combination wxth
heparin, a marked synergistic effect was observed. The
Factor IX mAbs had a modest effect on aPTT and even at
the highest dose, the increase in clotting time did not
exceed the 3-fold limit of standard anti-coagulant
practiced clinically. Most notably, the low dose of
Factor IX mAb BC2 in combination with ASA did not change
the aPTT.

in Fig 5, the data indicate that PT was also
significantly prolonged by heparin, at the two higher
doses, and by the ASA . heparin combination, but not by
any of the Factor IX mAb doses alone or in combination
with ASA.

42

P50438-1

The effect of heparin, ASA and Factor IX mAb on
carotid artery occlusion is shown in Fig. 6. The
results indicate that the carotid arteries of all of the
vehicle-treated animals occlude in response to the
injury. Heparin dose dependently inhibited the
occlusion of the carotid artery. At the highest dose,
heparin completely prevented the occlusion of the
carotid artery; at this dose however, no coagulation
could be initiated. ASA alone had only a minor effect
on carotid occlusion. ASA in combination with heparin
also failed to completely prevent carotid occlusion.
Factor IX mAb completely blocked carotid occlusion at
the two higher doses, which have not prolonged
coagulation beyond the clinically desired target. The
lower dose of Factor IX mAb, that largely failed to
secure patency alone, demonstrated complete inhibition
of carotid occlusion when administered in combination
with ASA.

The effect of heparin, ASA and Factor IX mAb on
thrombus weight is shown in Fig. 7. Heparin dose-
dependently reduced thrombus mass in the carotid
artery. However, some residual thrombus was still found
in the carotid artery in spite of complete blockade of
coagulation. ASA alone or in combination with heparin
(30 U/kg regimen) had only a partial effect on thrombus
weight. Factor IX mAb dose-dependent ly reduced thrombus
mass and the high dose virtually prevented completely
thrombus formation. Moreover, the combination of the
low dose anti-Factor IX mAb and ASA, a regimen that
completely prevented carotid occlusion without adversely
affecting the coagulation indices, completely prevented
thrombus formation.

The studies conducted in the rat carotid thrombosis
model clearly demonstrate the efficacy of Factor IX mAb
in prevention of thrombosis in a highly thrombogenic
arterial injury model. Most notably, the efficacy of

43

P50438-1

the Factor IX mAb was demonstrated within the desired
therapeutic anticoagulant target defined by the aPTT.
Furthermore, heparin, the current standard
anticoagulant, reached efficacy comparable to Factor IX
mAb only at doses that severely compromised coagulation
to the extent of producing non-coagulable blood,
interestingly, the observed potentiation and synergy
acquired by ASA joint treatment with heparin was also
demonstrated when ASA was given with anti-Factor IX mAb
However, unlike the combination of heparin and ASA which
resulted in potentiation of both the anti-thrombotic and
anti-coagulant effects, the combination of Factor IX mAb
and ASA resulted in potentiation of the anti-thrombotxc
efficacy with no consistent effect on ex vivo blood
coagulation parameters. Taken together, the data show a
superior antithrombotic capacity of Factor IX mAb
compared to heparin, ASA or a combination of heparin and

ASA.

Example 4

^-.^.^^^r. Micr of^roPY of R at Thrombosis Model

Segments of rat carotid artery were collected from
sham, ferric chloride only and ferric chloride -h 6 mg/kg
Factor IX antibody, 3/group, 15 minutes after
application of ferric chloride. The arteries were fixed
by perfusion with formaldehyde and ligated above and
below the lesioned area. Fixed arteries were
dehydrated, incubated in hexamethyldisilazane and dried
in a desiccator. Dried arteries were opened lengthwise,
placed on Scanning Electron Microscopy (SEM) stubs and
sputter coated with gold.

SEM of sham arteries revealed an essentially normal
endothelium with rare scattered platelets. There were a
few breaks in the endothelium, probably as a result of
mechanical damage during surgery and the underlying
basement membrane was covered by a carpet of platelets.

44

P50438-1

NO evidence of thrombus formation was observed in the

sham rats . , • j

SEM of the arteries treated with ferric chloride
revealed large mural thrombi which occupied a large
portion of the lumen of the vessel. The thrombi were
composed of aggregated platelets, red blood cells and
amorphous and fibrillar proteinaceous material. The
proteinaceous material is consistent with fibrin. The
endothelium of the arteries was mostly obscured by the
large thrombi, where visible, the endothelium overlyrng
the region treated with ferric chloride was covered by
numerous adherent platelets and amorphous proteinaceous

material . . ■ j

SEM of the arteries treated with ferric chlorxde
from rats also treated with Factor IX antibody, revealed
the lumen of the vessels to be largely free of thrombus.
The endothelium overlying the region treated with ferrxc
chloride showed extensive damage and some areas were
covered by adherent platelets and platelet aggregates
but there was little or no proteinaceous material.

Example 5
HtAb P^^ "-^^^ and T.ight-. Chain cDNA
ff^rpiQTinft Analysis

Total RNA was purified by using TriReagent
(Molecular Research Center, Inc., Cincinnati, OH)
according to the manufacturer's protocol. RNA was
precipitated with isopropanol and dissolved in 0.5% SDS
and adjusted to 0 . 5M NaCl . Poly A"^ RNA was isolated
with Dynabeads Oligo (dT)25 (^nal A.S., Lake Success,
NY) according to the manufacturer's protocol. Poly A
RNA was eluted from the beads and resuspended xn TE
buffer. Twelve aliquots of 100 ng of RNA were reverse
transcribed with a RT-PCR kit per the manufacturer's
instructions (Boehringer Mannheim Cat. No. 1483-188)
using a dT oligo for priming. For the heavy chaxn, PCR

45

P50438-1

amplifications of 6 RNA/DNA hybrids were carried out for
25 cycles using a murine IgG2a hinge primer (SEQ ID NO:

1) and a heavy chain signal sequence primer (SEQ ID NO:

2) . Similarly, for the light chain, PGR amplif icatons
of 6 RNA/DNA hybrids were carried out for 25 cycles
using a murine kappa primer (SEQ ID NO: 3) and a
degenerate light chain signal sequence primer (SEQ ID
NO- 4) The PGR products from each of the 12
amplifications were ligated in a PGR2000 vector (TA
cloning Kit, Invitrogen, Cat. No. K2000-01) . Golonxes
of recombinant clones were randomly picked and
minipreparations of plasmid DNA were prepared using an
alkaline extraction procedure described by Birnboim and
Doly in Nucl. Acids Res. 1, 1513 (1979). The isolated
plasmid DNA was digested with EcoRI and analyzed on a

0 8% agarose gel. Double-stranded cDNA inserts of the
appropriate size, i.e., -700 bp for the heavy chaxn and
-700 bp for the light chain, were sequenced by a
modification of the Sanger method. The sequence of all
12 of the heavy and light chains were compared to
generate a consensus BG2 heavy chain variable region
sequence (SEQ ID NO: 5) and consensus BC2 light chain
variable region sequence (SEQ ID NO: 6).

sequence analysis of the BC2 heavy chain variable
region cDNA revealed a 363 nucleotide open reading frame
encoding a 121 amino acid sequence (SEQ ID NO: 7) . The
heavy chain GDRl, 2 and 3 sequences are listed in SEQ ID
NOs- 8, 9 and 10, respectively.

sequence analysis of the BG2 light chain variable
region cDNA revealed a 321 nucleotide open reading frame
encoding a 107 amino acid sequence (SEQ ID NO: 11) • The
light chain GDRl, 2 and 3 sequences are listed in SEQ ID
NOs: 12, 13 and 14, respectively.

46

P50438-1

Eacamole 6
Tft,«.aTiigeti Antibodies

Six humanized antibodies designated SB 249413, SB
249415, SB 249416, SB249417, SB 257731 and SB 257732
were designed to contain the murine CDRs described above
in a human antibody framework.
SB 249413

SB 249413 contains the heavy chain F9HZHC 1-0 and
the light chain F9HZLC 1-0. The synthetic variable
region humanized heavy chain F9HZHC 1-0 was designed
using the first three framework regions of the heavy
chain obtained from immunoglobulin RF-TS3'CL (Capra,
J D et al., J. Clin. Invest. 86, 1320-1328 (1990)
identified in the Kabat database as Kabpro : HhclOw) and
the BC2 heavy chain CDRs described previously. No
framework amino acids substitutions which might
influence CDR presentation were made. Four overlapping
synthetic oligonucleotides were generated (SEQ ID NOs :
15 16, 17 and 18) which, when annealed and extended,
code for the amino acids representing the heavy chain
variable region through and including CDR3 (SEQ ID NOs:
19 and 20) . This synthetic gene was then amplified
using PCR primers (SEQ ID NOs: 21 and 22) and ligated
into the PCR2000 vector (TA cloning Kit, Invitrogen,
cat. NO. K2000-01) and isolated from a Spel, Kpnl
restriction digest. A second DNA fragment coding for
the campath signal sequence including the first five
amino acids of the variable region (SEQ ID NOs: 23 and
24) was made by PCR amplification of the appropriate
region of a construct encoding a humanized anti-
Respiratory Syncitial Virus heavy chain (SEQ ID NO: 25)
with two primers (SEQ ID NOs: 26 and 27) and digesting
with the restriction enzymes EcoRI and Spel . The two
fragments generated were ligated into an EcoRl, Kpnl
digested pFHZHC2-6pCD mammalian cell expression vector
which contained the remainder of a human consensus

47

P50438-1

framework 4 and IgGl constant region. The vector
contained a single amino acid mutation of the pFHZHC2-
3pCD vector described in published International Patent
Application NO. WO94/05690. The final residue of
framework 2 (residue 49) was mutated from Ser to Ala by
digesting pFHZHC2-3pCD with Xbal and EcoR5 and inserting
a linker generated from two synthetic oligonucleotides
(SEQ ID NOs: 28 and 29) . The sequence of the F9HZHC 1 0
insert is shown in SEQ ID NOs: 3 0 and 31.

The synthetic variable region humanized light chain
F9HZLC 1-0 was designed using the framework regions of
the human light chain obtained from immunoglobulin
LS8'CL (Carmack et al . , J. Exp. Med. 169, 1631-1643
(1989) identified in the Rabat database as
Kabpro:Hkl318) and the BC2 light chain CDRs described
previously. No framework amino acids substitutions
which might influence CDR presentation were made. Two
overlapping synthetic oligonucleotides were generated
(SEQ ID NOs: 32 and 33) which, when annealed and
extended, code for amino acids representing the light
chain variable region (SEQ ID NOs: 34 and 35) . This
synthetic gene was then amplified using PGR primers (SEQ
ID NOs: 36 and 37) and ligated into the pCR2000 vector
(TA cloning Kit, Invitrogen, Cat. No. K2000-01) , and
isolated from a Seal, 5acII restriction digest. A
second DNA fragment coding for the campath signal
sequence including the first two amino acids of the
variable region (SEQ ID NOs: 38 and 39) was made by PGR
amplification of the the appropriate region of a
construct encoding a humanized anti-Respiratory
Syncitial Virus heavy chain (SEQ ID NO: 25) with the two
primers (SEQ ID NOs: 26 and 40) and digesting with the
restriction enzymes EcoRl and Seal. The two fragments
generated were ligated into an EcoRl, SacII digested
pFHzLGl-2pCN mammalian cell expression vector which
contained the remainder of a human framework 4 and kappa

48

P50438-1

constant region. The vector contained a single amxno
acid mutation of the pFHZLCl-lpCN vector described xn
published international Patent Application No.
WO94/05690. A framework 2 residue was mutated from Ser
to Pro by digesting pFHZLCl-pCN with Smal and Kpnl and
inserting a linker generated from two synthetic
oligonucleotides (SEQ ID NOs: 41 and 42). The sequence
of the F9HZLC 1-0 insert is shown in SEQ ID NOs: 43 and

44.

SB 249415 ^ ^

SB 249415 contains the heavy chain F9HZHC 1-1 and
the light chain F9HZLC 1-1- These heavy and light chain
constructs are based on F9HZHC 1-0 and F9HZLC 1-0,
respectively, however, they have framework amino acid
substitutions which can influence CDR presentation.

F9HZHC 1-1 has three framework amino acid
substitutions which might influence CDR presentation.
TWO overlapping synthetic oligonucleotides were
generated (SEQ ID NOs: 45 and 46), which when annealed
and extended, code for amino acids representing the
altered portion of the heavy chain variable region
altered (SEQ ID NOs: 47 and 48). This synthetic gene
was then amplified using PGR primers (SEQ ID NOs: 49 and
50), ligated into the pCR2000 vector (TA cloning Kit,
invitrogen, Cat. No. K2000-01) and isolated from a
EcoNI, Kpnl restriction digest. This fragment was
ligated into EcoNI, Xpnl digested F9HZHC1-0 (SEQ ID NO:
30) vector. The sequence of the F9HZHC 1-1 insert is
shown in SEQ ID NOs: 51 and 52.

F9HZLC 1-1 has four framework amino acids
substitutions which can influence CDR presentation. TVo
synthetic oligonucleotides were generated (SEQ ID NOs:
53 and 54), which when annealed, have Kpnl and BaiiTHI
cohesive ends, and code for amino acids representing the
altered portion of the light chain variable region (SEQ
ID NO: 55). F9HZLC 1-0 (SEQ ID NO: 43) was digested

49

P50438-1

with the restriction enzymes Kpnl and BaMll and ligated
to the synthetic DNA. The sequence of the F9HZLC 1-1
insert is shovm in SEQ ID NOs : 56 and 57.
SB 249416

SB 249416 contains the heavy chain F9HZHC 1-1
(described above) (SEQ ID NO: 52) and the light chain
F9HZLC 1-2. The light chain construct is based on
F9HZLC 1-1, however, it has one additional framework
amino acid substitution which can influence CDR

presentation.

TWO synthetic oligonucleotides were generated (SEQ
ID NOs: 58 and 59), which when annealed, have BamHI and
Xbal cohesive ends and code for amino acids representxng
the altered portion of the light chain variable regxon

(SEQ ID NO: 60) . F9HZLC 1-1 (SEQ ID NO: 56) vector was
digested with the restriction enzymes BaMll and Xbal and

ligated to the synthetic DNA. The sequence of the

F9HZLC 1-2 insert is shown in SEQ ID NOs: 61 and 62.

.qg 249417

SB 249417 contains the heavy chain F9HZHC 1-1
(described above) (SEQ ID NO: 52) and the light chain
F9HZLC 2-0. A F9HZLC 2-0 synthetic variable region
humanized light chain was designed using the framework
regions of the human light chain obtained from
immunoglobulin REI (Palm and Hilschmann, Z. Physiol.
Chem 354. 1651-1654 (1973) identified in the Kabat
database as Kabpro : HKLlll) and the BC2 light chain CDRs
described previously. Five amino acid consensus human
substitutions were introduced. Six framework amino
acids murine substitutions which can influence CDR
presentation were made. Two overlapping synthetic
oligonucleotides were generated (SEQ ID NOs: 63 and 64)
Which, when annealed and extended, code for amino acids
representing the light chain variable region (SEQ ID
NOS- 65 and 66) . This synthetic gene was then amplified
using PGR primers (SEQ ID NOs: 67 and 68), ligated into

50

P50438-1

the PCR2000 vector (TA cloning Kit, Invitrogen, Cat. No.
K2000-01) and isolated from a Seal, SacII rastrictxon
digest A second DNA fragment coding for the campath
signal sequence including the first two amino acids of
the variable region (SEQ ID NO: 38) was made by PCR
amplification of the the appropriate region of a
construct encoding a humanized anti-Respiratory
syncitial Virus heavy chain (SEQ ID NO: 25) with two
primers (SEQ ID NOs : 26 and 69) and digesting wrth the
restriction enzymes EcoRI and Seal. A third DNA
fragment encoding the remainder of a human framework 4
(SEQ ID NO: 70) and having SacII and Narl cohesive ends
was generated by annealing two synthetic
oligonucleotides (SEQ ID NOs: 71 and 72). F9HZLC 1-0
(SEQ ID NO: 43) was digested with the restriction
enzymes EcoRI and Narl and ligated to the three DNA
fragments. The sequence of the F9HZLC 2-0 insert is
shown in SEQ ID NOs: 73 and 74.
SB 257731

iri57731 contains the heavy chain F9HZHC 1-1 (SEQ

ID NO: 52) and the light chain F9HZLC 1-3, a single
axnino acid mutation of F9HZLC 1-2 (SEQ ID NO: 62).
F9HZLC 1-2 was PCR amplified with two primers (SEQ ID
NOS- 26 and 69) and digested with the restriction
enzymes EcoRI and Seal. A 94 bp fragment (SEQ ID NOs:
75 and 76) was isolated. The fragment was Ixgated xnto
EcoRI, seal digested F9HZLC 1-2 vector to produce the
light chain construct F9HZLC 1-3. The sequence of the
F9HZLC 1-3 insert is shown in SEQ ID NOs: 77 and 78.
SB 257732

ir^57732 contains the synthetic variable regxon
humanized heavy chain F9HZHC 3-0 and light chain F9HZLC
3-0 Four overlapping synthetic oligonucleotides were
generated (SEQ ID NOs: 79, 80, 81 and 82) which, when
annealed and extended, code for the amino acids
representing the heavy chain variable region being

51

P50438-1

altered (SEQ ID NOs: 83 and 84). This synthetic gene
was then amplified using PGR primers (SEQ ID NOs: 85 and
86) ligated into the pCR2000 vector (TA cloning Kit,
invitrogen. Cat. No. K2000-01) and isolated from a 5tul,
Kpnl restriction digest. The isolated fragment was
ligated into Stul , Kpnl digested F9HZHC1-1 (SEQ ID NO:
52) vector. This vector was then digested with EcoRI,
Spel to remove the signal sequence. A DNA fragment
coding for the campath signal sequence (SEQ ID NO: 23)
including the first five amino acids of the variable
region was made by PGR amplification of F9HZHG1-0 with
two primers (SEQ ID NOs: 2 6 and 87) and digesting wxth
the restriction enzymes EcoRI and Spel . The fragment
generated was ligated into the vector. The sequence of
the F9HZHG3-0 insert is shown in SEQ ID NOs: 88 and 89.

Four overlapping synthetic oligonucleotides were
generated (SEQ ID NOs: 90, 91, 92 and 93) which, when
annealed and extended, code for amino acids representing
the light chain variable region (SEQ ID NOs: 94 and 95).
This synthetic gene was then amplified using PGR primers
(SEQ ID NOs: 96 and 97) and ligated into the pGR2000
vector (TA cloning Kit, Invitrogen, Gat. No. K2000-01) ,
and isolated from a 5cal, I^arl restriction digest. The
isolated fragment was ligated into Seal, Narl digested
F9HZLC1-3 (SEQ ID NO: 77) vector. The sequence of the
F9HZLC3-0 insert is shown in SEQ ID NOs: 98 and 99.

The humanized anti-Factor IX mAbs were expressed xn
CHO cells. A DG-44 cell line adapted for suspension
growth in serum-free medium was grown in 100ml of
protein-free medium containing IX nucleosides and 0.05%
F68 in 250 ml disposable sterile erlenmeyer flasks
(Corning) on a Innova 2100 platform shaker (New
Brunswick Scientific) at 150 rpm at 37°C in a 5% GO,, 95%
air humidified incubator. These cells were passaged at
4 X 10' cells/ml twice weekly. 15 ug each of the pCN-Lc-
Light Chain and pGD-Hc-heavy chain vectors were

52

P50438-1

linearized by digestion with i^otl, co-precipitated under
sterile conditions and resuspended in 50ul of IX TE
buffer (lOmM Tris, linM EDTA, pH 7.5) . The DNA was
electroporated using a Bio-Rad Gene Pulser (Bio-Rad
Laboratories) into the Acc-098 cells using the technique
of Hensley at al . in J. Biol. Chem. 269. 23949-23958
(1994) . 1.2 X 10' cells were washed once in 12.5 ml of
ice cold PBSucrose (PBS, 272mM sucrose, 7mM sodium
phosphate pH 7.4, ImM MgCl,) , resuspended in 0.8 ml of
PBS added to 50ul of the DNA solution and incubated on
ice for 15 min. The cells were pulsed at 380 V and 25
microfarads, then incubated on ice for 10 min. Cells
were plated into 96 well culture plates at 5 X 10
cells/plate in maintenance medium for 24 hr prior to
selection. Cells were selected for resistance to
400ug/ml G418 (Geneticin, Life Technologies, Inc.) in
maintenance medium. 24 hr prior to assay, the cells
were fed with 150ul of the maintenance medium.

conditioned medium from individual colonies was
assayed using an electrochemiluminescence (ECL)
detection method on an Origen analyzer (IGEN, Inc . ) .
5ee Yang et al . , Biotechnology, 12. 193-194 (1994).

All solutions necessary for the performance of the
assays {assay buffer) and for the operation of the
analyzer (cell cleaner) were obtained from IGEN. The
antibodies (anti-human IgG (g-chain specific) , Sigma
Chemicals and F(ab')3 Fragment to Human IgG (H+L) ,
Kirkegaard & Perry Laboratories Inc.) were labelled with
TAG-NHS-ester (IGEN, Inc.) at a 7 : 1 molar ratio of
TAG:protein, while the Protein A (Sigma) was labelled
with Biotin-LC-Sulfo-NHS-ester (IGEN, Inc.) at a 20:1
molar ratio Biotin :protein, both according to IGEN's
recommendations. Streptavidin-coated magnetic beads (M-
280) were obtained from Dynal .

immunoassays were performed using the following
protocol: per sample, 50ul of the streptavidin-coated

53

P50438-1

beads (final concentration 600 ug/ml diluted in PBS,
pH7.8, with 1.25% Tween) were mixed with 50ul Biotin-
Protein A (final concentration lug/diluted in PBS,
pH7.8, with 1.25% Tween) and incubated at room
temperature for 15min with agitation, 50ul of the TAG
antibodies (a mixture with a final concentration of 1.25
ug/ml F(ab'), Fragment to Human IgG (H+L) and 0.25 ug/ml
Anti-Human IgG (g-chain specific) diluted in PBS, pH7.8,
with 1.25% Tween) were added, the solution was then
added to 50ul of conditioned medium and incubated with
agitation at room temperature for 1 hr. 200ul of assay
buffer was added to the reaction mix and the sample
analyzed on the Origen I analyzer to measure ECL. The
results indicated that approximately 20-37% of the
colonies assayed secrete over 15 ng/ml of the antibody
with an average expression of about 150 ng/ml.

Humanized anti-Factor IX mAbs were purified from
the conditioned media using a Procep A capture step
followed by ion-exchange chromatography to reduce the
DNA burden. Procep A sorbent material (Bioprocessing
Ltd., Durham, England) was used to prepare a column with
a 1:1 diameter to height ratio. Clarified conditioned
media was loaded onto the column at about 150 cm/hr.
The column was washed sequentially with phosphate
buffered saline (PBS), PBS containing 1 M NaCl, and
finally with PBS. The bound material was recovered with
0.1 M acetic acid elution. The eluate was adjusted to
pH 5.5 and was diluted (1:4) with water. The diluted
solution was loaded onto an S-Sepharose column (2.5 x 13
cm) which was pre-equilibrated with 2 0 mM sodium
acetate, pH 5.5 at 80 cm/hr. The column was washed with
the acetate buffer until a steady baseline was obtained
and the bound protein was eluted with 20 mM sodium
phosphate, pH 7.4 at 25 cm/hr. The eluted material was
filtered with a 0.4 micron membrane and stored at 4^0.

54

P50438-1

Example 7
Mouae-Hum&n Chimer ic Antibody

100 ng of BC2 RNA were reverse transcribed with a
RT-PCR kit per the manufacturer's instructions
(Boehringer Mannheim Cat. No. 1483-188) using a dT oligo
for priming, and PGR amplified with synthetic Seal (SEQ
ID NO: 100) and Narl {SEQ ID NO: 101) primers to produce
the BC2 light chain variable region with Seal, Narl ends
(SEQ ID NOs: 102 and 103) . This DNA was ligated into
Seal, Warl digested F9HZHC1-3 (SEQ ID 77) and digested
with Seal, Narl to produce a mouse-human chimeric light
chain F9CHLC (SEQ ID NOs: 104 and 105) .

100 ng of BC2 RNA were reverse transcribed with a
RT-PCR kit per the manufacturer's instructions
(Boehringer Mannheim Cat. No. 1483-188) using a dT oligo
for priming, and PCR amplified with synthetic Spel (SEQ
ID NO: 106) and Nhel (SEQ ID NO: 107) primers to produce
the BC2 heavy chain variable region with Spel, Nhel ends
(SEQ ID NOs: 108 and 109). The campath signal sequence
was PCR amplified from the RSVHZ19 heavy chain (SEQ ID
NO: 25) with EcoRI (SEQ ID 26) and Spel (SEQ ID 87)
primers. These two DNA fragments were ligated into a
BcoRI, Nhel digested lL4CHHCpcd vector described in
published International Patent Application No.
WO95/07301, replacing the IL4 variable region with the
BC2 Factor IX mouse variable region, to produce a mouse-
human chimeric heavy chain F9CHHC (SEQ ID Nos : 110 and
111) -

Co-transfection and purification of the mouse-human
chimeric antibody chaFIX was accomplished as described
above for the humanized constructs.

55

P50438-1

Exam.pl 8

m^r^rj^^^Y >i»ni« «^ Facto ^ IX mAbs in Rat Thrombus
Model

In order to evaluate the efficacy of humanized
anti-Factor IX antibodies in prevention of arterial
thrombosis, the rat carotid artery thrombosis model as
described above in Example 3 was used. Baseline
parameters were established for carotid blood flow,
arterial pressure, heart rate, vessel patency and
activated partial thromboplastin time (aPTT) . Fifteen
minutes thereafter, carotid injury was effected for 10
minutes. The parameters were determined 60 minutes
after onset of carotid injury. Carotid thrombus was
also extracted from the carotid artery and weighed.

All agents were administered intravenously 15
minutes before the onset of carotid injury. The
following treatments were examined and compared to the
anti-Factor IX mAb BC2 .

1 . Vehicle

2. chaFIX: 3 mg/kg bolus

3. SB 249413: 3 mg/kg bolus

4. SB 249415: 3 mg/kg bolus

5. SB 249416: 3 mg/kg bolus

6. SB 249417: 3 mg/kg bolus

7. SB 257731: 3 mg/kg bolus

8. Heparin: 60 units /kg bolus + 2 units /kg/min
infusion

The aPTT was used as the primary criterion for
evaluation of efficacy versus bleeding liabilities of
the anti-coagulant/ thrombotic agents used in the study.
The results in Fig. 8 demonstrate that the humanized
Factor IX mAbs SB 249413, SB 249415, SB 249416, SB
249417 and SB 257731 had a modest effect on aPTT at 3 . 0
mg/kg which is within the clinical accepted range.

56

P50438-1

The effect of the Factor IX mAbs on thrombus mass
is shown in Fig. 9. The results indicate that all of
the humanized mAbs are equally effective in reducing
thrombus mass .

The studies conducted in the rat carotid thrombosis
model clearly demonstrate the efficacy of the humanized
Factor IX mAbs in prevention of thrombosis in a highly
thrombogenic arterial injury model. Most notably, the
efficacy of all of the humanized Factor IX mAbs was
demonstrated within the desired therapeutic
anticoagulant target defined by the aPTT.

Exeunple 9

Antibody Biochemical and B iophysical Properties

The molecular mass of SB 249417 was determined by
MALD-MS to be 148,000Da. Analytical ultracentrif ugation
of SB 249417 gave an identical value. In the presence
of Factor IX plus Ca2+, the antibodies derived from BC 2
sedimented with a mass of 248,000Da corresponding to the
combined mass of the mAb and two molecules of Factor IX.
No evidence of higher ordered aggregates was observed in
the presence or absence of Factor IX.

The kinetics of Factor IX binding to SB 249417 was
assessed by BIAcore analysis with antibody bound to an
immobilized protein A surface. Recombinant human Factor
IX (rhFIX, Genetics Institute) at 49 nM was used and
measurements performed in the presence of 5 mM Ca2+.
The interaction was characterized by rapid association,
kass = 2.0 X 105 M"! s"! and relatively slow off-rate,
kdiss = 4.1 X 10-4 s-1. The calculated for Factor IX
binding was 1.9 nM.

Table 1 summarizes the biophysical properties of SB
249417 .

57

P50438-1

Table 1

Summary of the Biophysical Properties of SB 249417

Isotype IgGl, kappa

Purity by SDS-PAGE >95% (under reducing

conditions)

Molecular Weight

Mass Spectrometry 148,000 Da

Analytical Ultracentrifugation 148,000 Da

Stoichiometry of Factor IX Binding

Isothermal Titration Calorimetry 1.5 moles Factor
IX: 1 mole mAb

Factor IX Binding Affinity

Isothermal Titration Calorimetry Kd= 4 nM at 25°C
Biosensor Kd= 2 nM

Factor IX Binding Kinetics

Biosensor k^ss = 2.0 x 10^ M"! s'l

kdiss =4 X 10-4 s-1

Table 2 summarizes the factor IX binding properties
of mAbs of the present invention. The calculated
dissociation constants were essentially identical within
experimental error.

P50438-1

Table 2

Kinetics of Factor IX Binding to Anti-Factor IX mAbs

249417

calc

. Kp <nM)

SB

2 .

0 xl05

4.1x10-4

1 . 9

BC2

4

.SxlO^

9.

1x10-4

1 . 9

Chf9

2.

4 xl05

3 .

.0x10-4

1 . 3

SB

249413

6 .

5 xl05

2 .

.8x10-3

3 .

.7-5.1

SB

249415

7 .

5 xl05

1,

.8x10-4

1

.1-2.3

SB

249416

5.

.2 xl05

4

.1x10-4

0.8

SB

257731

9 .

,2 xl05

9

.9x10-4

1.1

SB

257732

1,

.1 xl06

1

.2x10-3

1.5

The interactions between rhFIX and SB 249417 , BC2
and other hiimanized constructs were characterized by
titration microcalorimetry, which measures binding
interactions in solution from the intrinsic heat of
binding. Nine injections of 106 uM FIX were made into
the calorimeter containing 2 uM. mAb SB 249417. Binding
was detected in the first 4 injections as exothermic
heats. At the last 5 injections the mAb binding sites
were saturated with FIX and only background heats of
mixing were observed. The results indicated that the
equivalence point occurred at a molar binding ratio near
2 FIX per mAb, as expected. Nonlinear least squares
analysis of the data yield the binding affinity.

The rhFIX affinities of the mAbs were measured over
a range of temperature from 3 4-44°C in lOmM HEPES, lOmM
CaCl2, 150mM NaCl, pH 7.4. These data allow the
affinity at 37°C to be determined directly and the
affinity at 25°C to be calculated from the van't Hoff

59

P50438-1

equation. The data in Table 3 indicate that the
affinities of SB 249417, BC2 and its other humanized
constructs are within error (a factor of 2) the same.

Table 3

Titration Calorimetry Results for Anti-PIX loAbs

Kd, nM at 25°C Kd, nM at 37°C

Molar Binding
Ratio

BC2

SB
249413

SB
249415

SB
249417

SB
257732

20
12

1.4
1.9

The mAbs SB 249413, SB 249415, SB 249417 and SB
257732 all exhibited very similar thermal stabilities by
differential scanning calorimetry. Their unfolding Tms
ranged from 70-75°C indicating high stability against
thermally induced denaturation .

Example 10

Mechanism of Antibody-Mediated I nhibition of Factor IX

A library of chimeric constructs composed of
sequences of Factor IX spliced into the framework of the
homologous protein Factor VII was constructed and used
to map the epitope for the Factor IX BC2 mAb. See
Cheung et al . , Thromb. Res. 80, 419-427 (1995). Binding
was measured using a BiaCore 2000 surface plasmon

60

P50438-1

resonance device. The BC2 antibody was coupled directly
to the chip using the NHS/EDC reaction. Binding was
measured by 2 min of contact time at 20uL/min with 200
nM of each of the given constructs in 25 mM MOPS, pH
7.4, 0.15 M NaCl, 5 mM CaCl2 • Dissociation was
monitored for 3 min using the same buffer with no
protein. No binding was detected to the wild type
construct in the presence of 50 mM EDTA. The data are
presented in Table 4 .

Table 4

Summmary of Binding of Factor IX Constructs to BC2
Antibody

Construct

Degree of Binding

Plasma IXa

Binds

r-IX

Binds

Plasma VII

No Binding

IX LC/VII HC

Binds

IX-A/VII

Binds

VII gla/IX

No Binding

VII-A/IX

No Binding

VII

gla {IX 3-11) /IX

Binds

VII

gla (IX 3-6) /IX

Very Low Binding

VII

gla (IX 9-11) /IX

Very Low Binding

IX K5A

Binds

These data indicate that the constructs containing
the Factor IX light chain and Factor VII heavy chain (IX
LC/VII HC) ; the Factor IX gla and aromatic stack domains
(IX-A/VII); residues 3-11 of Factor IX gla domain within

61

P50438-1

the Factor VII gla domain (VII gla (IX 3-11) /IX) ; and
Factor IX having a lysine to alanine substitution at
residue 5 (IX K5A) exhibit binding to BC2 . The VII gla
{IX 3-11) /IX construct exhibited BC2 binding equivalent
to wild type Factor IX (plasma IXa and r-IX) . Thus, the
BC2 antibody binds to an epitope contained within
residues 3-11 of the Factor IX gla domain.

Example 11

ni^^^i-m^TYt of Arte r n^l Thrombosis with Anti-Factor IX
Antibody and Tisgue Plasm inogen Activator

Administration of tPA with or without adjunctive
therapies, was initiated following complete occlusion of
the carotid artery. Blood flow in the artery was
continuously monitored.

Male Sprague-Dawley rats (Charles River, Raleigh, NC)
weighing 3 00-490 gm were anesthetized with sodium
pentobarbital (55 mg/kg, i.p.). The rats were placed
dorsal on a heated (37°C) surgical board and an incision
was made in the neck; the trachea was isolated and
cannulated with a PE-240, Intramedic tube. The left
carotid artery and jugular vein were then isolated. A
Parafilm M sheet (4 mm2 , American National Can) was
placed under the carotid artery and an electromagnetic
blood flow probe (Carolina Medical) was placed on the
artery to measure blood flow. A cannula (Tygon, 0.02" x
0.04", Norton Performance Plastics) was inserted into
the jugular vein for drug administration. The left
femoral artery was then isolated and cannulated for
measurement of blood pressure and collection of blood
samples .

Thrombosis in the carotid artery was initiated with
a 6.5 mm diameter circular patch of glass micro-filter
paper saturated with FeCls solution (50%) placed on the
carotid artery downstream from the flow probe for 10
minutes as described in Example 3 . In this well-

62

P50438-1

characterized model, thrombus formation is usually
complete within 15 min.

The anti-Factor IX antibody, SB 249415 was
administered as a bolus in combination with tPA
(Genentech, South San Francisco, CA) , while heparin
(Elkins-Sinn Inc., Cherry Hill, NJ)was administered as a
bolus followed by infusion. All drug infusions
continued to the end of the experimental period - 60
minutes from the time of vessel occlusion. Blood
samples, 1 mL, were collected for aPTT and FT assay at
0, 30 and 60 min (end of study) from the femoral artery
into 3.8% citrate solution and centrifuged. aPTT and
prothrombin time (PT) were monitored by a fibrometer
(BBIL, Baxter Dade or MLA Electra 800 Automatic
coagulation Timer) with standard procedures. At the end
of the experiment, the thrombus was extracted from the
carotid artery and weighed.

All data are presented as mean group values + SEM
for the indicated number of rats in each group. ANOVA
and Bonferoni tests for multiple comparisons were used
for between group analyses and a p < 0.05 accepted as
significance.

Formation of an occlusive thrombus occurs
approximately 15 min after initiation of arterial injury
by application of the FeCls treated patch to the rat
carotid artery. As shown in Fig. 10, with tPA alone,
reperfusion of the occluded vessel was only observed
following administration of a dose of 9 mg/kg tPA with
67% of the treated vessels exhibiting regain of blood
flow during the 60 min protocol. At this dose of tPA
inclusion of 60 U/kg heparin or 3 mg/kg anti-Factor IX
antibody, SB 249415, did not result in a further
increase in the incidence of reperfusion suggesting that
in the FeCl3 injury model about 30% of the thrombi are
refractory to lysis.

63

P50438-1

The results in Fig. 10 indicate that, at lower
doses of tPA, the incidence of reperfusion is
significantly dependent upon which anticoagulant was co-
administered with the thrombolytic. When 60 U/kg
heparin was administered the percentage of vessels
showing reperfusion decreased dramatically with only
12.5% and 40% reperfusion observed with 3 and 6 mg/kg
tPA, repectively. Co-administration of 3 mg/kg SB
249415 with the tPA, however, achieved greater than 60%
reperfusion with 3 mg/kg tPA and 79% reperfusion with
the 6 mg/kg tPA dose. Thus, the anti-Factor IX antibody
significantly shifts the thrombolytic dose response
curve allowing reperfusion with lower doses of
thrombolytic agent.

Thrombolysis and clot formation are dynamic
processes and periods of patency followed by re-
occlusion were sometimes observed. Since carotid blood
flow was monitored continuously during the 60 min
experimetal protocol, it was also possible to quantitate
the total time of carotid patency. As shown in Fig. 11,
the total period of vessel patency is substantially
increased by combination of 3 mg/kg anti-Factor IX
antibody plus tPA. This is particularly evident at the
lowest and the intermediate doses of tPA, 3 and 6 mg/kg,
respectively. At a combined dose of 3 mg/kg SB 249415
plus 3 mg/kg tPA, the total patency time was 30.6 ± 9.2
min compared to 7 . 1 ± 7 . 1 min f or the combination of 60
U/kg heparin plus 3 mg/kg tPA. Patency time was zero
with 3 mg/kg tPA alone. With a dose of 6 mg/kg tPA, co-
administration with heparin increases patency time only
slightly to 12.9 ± 6.0 min whereas the tPA-SB 249415
combination achieves maximal patency time of 3 8.7 ± 8.4
min. Only at the highest dose of tPA (9 mg/kg) does the
heparin combination approach the patency achieved with
SB 249415, 31.9 ± 4.8 min and 38.0 ± 8.4 min,
respectively .

64

P50438-1

Rapid restoration of blood flow following arterial
infarct is critical to minimizing damage to the ischemic
tissue. The results in Fig. 12 indicate that the
combination of anti-factor IX antibody with tPA resulted
in decreased time to reperfusion compared to tPA alone
or heparin plus tPA and that this is achieved with lower
doses of tPA. When thrombolysis was effected with 3
mg/kg SB 249415 plus 3 mg/kg tPA the time to
thrombolysis was 29.4 ± 9.2 min. With 3 mg/kg tPA,
alone, no reperfusion was observed. With 60 U/kg
heparin plus 3 mg/kg tPA the time to thrombolysis was
52.8 ± 7.1 min. At higher doses of tPA, 6 and 9 mg/kg,
the antibody plus tPA treatment regimen achieved initial
thrombolysis in 19.4 ± 6.3 and 20.8 ± 8.7 min,
respectively. In the absence of added anticoagulant,
the time to thrombolysis was 60 min ( i.e., the limit of
the experimental protocol) and 27.5 ± 6.4 min for doses
of 6 and 9 mg/kg, respectively. With addition of 60
U/kg heparin, the corresponding times to thrombolysis
were 44.0 ± 7.1 and 27.0 ± 4.9 min. Thus, earlier
reperfusion was always achieved with SB 249415 than with
heparin or with tPA alone.

Eagamole 12

Effect of AT^ti-Facto r- TX Antibody on Hemosf.atlc Function

The impact of anti-factor IX or heparin as
adjunctive agents on the maintanence of hemostatic
function was determined by monitoring levels of
fibrinogen, plasminogen and alpha-2-antiplasmin at the
end of the treatment period in rats treated with tPA
alone, tPA plus heparin and tPA plus SB 249415 and the
results were compared to vehicle treated animals. As
shown in Fig. 13, increasing doses of tPA resulted in
decreased levels of each of the hemostatic markers
measured. Alpha-2-antiplasmin levels dropped from about
90% in animals not treated with tPA to about 20% as the

65

P50438-1

dose of tPA was increased to 9 mg/kg. Plasminogen
levels dropped from an average of about 100% without tPA
treatment to about 40% in the 9 mg/kg treatment group.
Likewise, fibrinogen levels dropped from about 150 mg/dL
to about 90 mg/dL in the high-dose tPA group,
interestingly, the selection of the adjunctive agent
does not appear to significantly effect any of these
markers. At each tPA dose, similar levels of alpha-2-
antiplasmin, plasminogen and fibrinogen were observed xn
animals given vehicle, 30 or 60 U/kg heparin or 1 or 3 ^
mg/kg SB 249415; the observed decrease in the hemostatic
markers is only a function of dose of the thrombolytic
agent, tPA, and these decreases, especially in the case
of fibrinogen, appear to be particularly large with tPA
doses greater than 6 mg/kg, i.e., in the 9 mg/kg hxgh-
dose group.

The effects of the different treatment regimens on
the standard aPTT coagulation assay aPTT was also
monitored (Fig. 14). With increasing doses of tPA the
aPTT increased from 19.3s ± 0.6s to 30.0s ±1.6s for
vehicle and 9 mg/kg tPA, respectively. Administration
of 3 mg/kg SB 249415 produced a limited increase m the
aPTT of control animals to 49 . 6 s ± 6 . 4 s . When SB
249415 was co-administered with tPA the observed
increase was slightly larger and was dependent upon the
dose of tPA. combination of SB 249415 with 3 mg/kg tPA
produced an aPTT of 58.3 s ±5.2 s whereas combination of
SB 249415 with the 9 mg/kg dose of tPA increased the
aPTT to 77.3 s ± 19.7 s. Administration of either the
3 0 U/kg or 60 U/kg dose of heparin resulted in large
increases in the aPTT. Without tPA the aPTT ranged from
about 300 s to 600 s for 30 and 60 U/kg doses,
respectively. In tPA treated animals the aPTT was about
800 s.

The elevation of the aPTT obtained with heparin,
particularly when coupled with the perturbation of

66

P50438-1

hemostatic parameters due to the need for high doses of
tPA to achieve effective reperfusion, is likely to
contribute to bleeding liabilities. Conversely, SB
249415 does not cause major elevation of the aPTT and
enables the use of lower doses of tPA providing
significant advantage in thrombolytic therapy in
myocardial infarction and stroke.

The present invention may be embodied in other
specific forms without departing from the spirit or
essential attributes thereof, and, accordingly,
reference should be made to the appended claims, rather
than to the foregoing specification, as indicating the
scope of the invention.

67